# Line 21 > Line 21 is a UK-based live captioning platform and agency. It delivers live captions, translations, and AI dubbing for events, streams, venues, broadcasts, and meetings, combining professional human captioners with AI speech recognition, an AI Proofreader, machine translation, and speech-to-speech dubbing. - Languages: 120; speech recognition in 104, AI translation in 93, AI dubbing in 75. - Plans: Pay as you go ($0/month), Essential ($199/month), Pro ($599/month), Enterprise (tailored). - Delivery: browser link, on-screen overlay, Zoom / Teams / YouTube via API token, CEA-608/708 in RTMP, HLS streaming with multi-language captions. - Contact: hello@line-21.com or https://line-21.com/contact. Product app: https://app.line-21.com. Every page listed here is Markdown. Any page on the site is also available as Markdown by adding `.md` to its URL (`https://line-21.com/index.md` for the homepage) or by requesting it with `Accept: text/markdown`. Structured data: https://line-21.com/languages.json and https://line-21.com/pricing.json. Everything in one file: https://line-21.com/llms-full.txt. --- # Line 21 — Every voice, every language. Live Live captions, translations, and AI dubbing for streams, venues, and broadcast. Run it on our platform — or hand it to our team. Line 21 is a UK-based live captioning platform built on our own technology: an AI Proofreader that checks every caption against your event's names and jargon, and a cloud video encoder that puts captions into any stream. Around them we orchestrate 15+ AI models and professional human captioners, to deliver accurate captions, translations and AI dubbing in 100+ languages, live. ## Workflows ### Human captioning CART writers work the way they already work: your machine, your CAT software, our delivery. The platform is free for individual captioners and the agencies that book them. - Latency: 0.1s - Accuracy: 98% [More about Human captioning](https://line-21.com/services/human-live-captions) ### AI Captions & Translations RTMP, SRT, raw audio, or a bot in the meeting. 100+ languages. Multiple AI models behind one endpoint, because no single vendor wins on every language. - Latency: 3s - Accuracy: 95% - Languages: 100+ [More about AI Captions & Translations](https://line-21.com/features/speech-recognition) ### Video Encoding Custom cloud video encoder: CEA-608/708 into RTMP for broadcast and YouTube, multi-track SRT for several languages in one stream, HLS tuned for scale. Stream anywhere you already stream. - Latency: 5s [More about Video Encoding](https://line-21.com/features/caption-encoding) ### AI Dubbing Speech-to-speech, so the voice never round-trips through text. Real-time, low latency, priced for events that prefer not to fund an interpreter booth. - Latency: 10s [More about AI Dubbing](https://line-21.com/features/ai-dubbing) ## AI Proofreader An industry-first safeguard that learns the context of your event and corrects live captions before the audience sees them — names, jargon, and the line that was never meant to be funny. [How the AI Proofreader works](https://line-21.com/features/ai-proofreader) ## Destinations ### Browser link A URL or a QR on the seat. Captions, translations, and dubbed audio on any phone — no app, no login. [More about Browser link](https://line-21.com/features/integrations) ### Overlay Open captions over IMAG and in-room screens. Place the app on the picture, style the type, leave it there. [More about Overlay](https://line-21.com/features/integrations) ### YouTube & destinations CEA-608/708 in the stream you already send. YouTube, Twitch, Facebook, Vimeo, broadcast — captions live in the player's CC button. [More about YouTube & destinations](https://line-21.com/features/caption-encoding) ### HLS streaming We package the stream, insert every language, and host it on a global CDN. Viewers pick a language as closed captions — HLS is the format that actually allows that. - Multi-language closed captions - AI Dubbing - Global CDN - Our player or yours [More about HLS streaming](https://line-21.com/features/hls-streaming) ## What customers say Rated 4.0 out of 5 from 4 reviews on [Trustpilot](https://www.trustpilot.com/review/line-21.com). > **For several years we have been searching…** — Nils Bryntesson, SE, 5 October 2026 > For several years we have been searching for a solution where we could do closed captions, using the broadcast standard CEA 608-708 with human captioners, to Youtube Live. Line-21 has been the first service provider, out of quite a few, that was able to reliably make this work. > The staff has been incredibly service minded and agile when problems has arisen or when we’ve had requests out of the ordinary. We are also in joint development for new services that are required in our local market. Thus far we are very happy with our cooperation! > **Exceptional Partner for Our Business** — Darian Ghassemi, US, 2 December 2025 > We've had an outstanding experience working with Line 21. From day one, they've demonstrated professionalism and delivered a service that clearly puts the customer first. > What really sets them apart is their team. Paul and Fabrizio are incredibly communicative, transparent, and open throughout every stage of our collaboration. They're responsive, proactive, and make it easy to stay aligned on goals and deliverables. > Line 21 has proven to be more than just a client; they're a true partner. They've made collaboration simple and effective, which is exactly what you want when working with an external team. I'd highly recommend them to anyone looking for a reliable and professional partner. > **Nous avons travaillé avec Line21 dans…** — Nathalie MOREL, FR, 25 November 2025 > Nous avons travaillé avec Line21 dans le cadre du sous-titrage multilingue d'un événement interne. Toute l'équipe s'est montrée disponible et réactive et l'événement fût un grand succès grâce à la qualité du sous-titrage. Je ne peux que recommander Line21! > **Solid systm, solid support** — Jody Botting, GB, 14 November 2025 > The level of support and approachable attitudes of the team fills me with confidence every time. Not only am I using a product that excels, but there are friendly faces there to guide you if you need it - thank you Line 21! ## Community - **Discord**: This is our home. Here you can chat with us, get latest updates, real-time support, etc. - **Docs**: This is our knowledge center. Here you find out how to use Line 21, from general use cases down to specifics. - **Social Media**: This is where the fun is. We are present on several social media networks where we cover Line 21 and industry updates. Discord: https://discord.gg/ZgCYyXywCE · LinkedIn: https://www.linkedin.com/company/line-21-captions/ · YouTube: https://www.youtube.com/@ccbyline21 --- # How we use AI AI runs through every stage of a live caption, from the first word heard to the last voice dubbed. Each step is a service you can switch on, tune or leave to a human. ## 1. Speech recognition Speech-to-text models turn live audio into captions. Line 21 recommends the best engine for each language and lets you switch; some engines also tell speakers apart and follow conversations that change language. - Engines: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe - Languages: 104 - Latency: 3s [Speech recognition docs](https://line-21.com/docs/methods/asr) ## 2. AI Context Give the AI the briefing a captioner would get: scripts, slides, agendas, and names. The AI Context Builder turns it into a description, names, and specialised language, which feed the proofreader and give speech engines a dictionary of words to expect. - Engines: OpenAI, Anthropic [AI Context docs](https://line-21.com/docs/ai-services/ai-context) ## 3. AI Proofreader A language model checks every caption line against that context, fixing names, brands, punctuation, and filler words, and it can finish before the audience sees a word. It works on captions from people and machines alike. - Engines: OpenAI, Anthropic - Latency: <1s [How the AI Proofreader works](https://line-21.com/features/ai-proofreader) · [AI Proofreader docs](https://line-21.com/docs/ai-services/ai-proofreader) ## 4. AI translation Every caption line is translated live into as many languages as you need, with the engine chosen per language and custom dictionaries for brand names and terms. - Engines: DeepL, Google, Azure, OpenAI - Languages: 93 [AI translation docs](https://line-21.com/docs/ai-services/ai-translations) ## 5. AI dubbing Speech-to-speech interpretation: the voice is carried straight into another language's audio track, without a round trip through text, for audiences who would rather listen than read. - Engines: Gemini Live Translate - Languages: 75 - Latency: 10s [How AI dubbing works](https://line-21.com/features/ai-dubbing) · [AI dubbing docs](https://line-21.com/docs/ai-services/ai-dubbing) ## 6. AI call agent A bot joins your Zoom call and transcribes it live, translating as it goes. Audio is used in the moment and never stored. Microsoft Teams and Google Meet are in development. - Runs on the speech recognition engines above. [AI call agent docs](https://line-21.com/docs/ai-services/ai-call-agent) ## People stay in control AI is a set of tools in the workflow, not a replacement for the people who run it. - **Human captioners, at no cost**: Bring your own captioners, respeakers, or Velotype writers; delivering their captions is free. ([Pricing](https://line-21.com/pricing)) - **Your choice of engine**: We recommend an engine per language, and you can test and switch it. ([Speech recognition docs](https://line-21.com/docs/methods/asr)) - **Checked before it airs**: Hold captions until the proofreader is done, so the audience only sees the final text. ([Before or after distribution](https://line-21.com/docs/ai-services/ai-proofreader#before-or-after)) - **Context you can edit live**: Change names and terms during the event; the proofreader uses them from the next line. ([AI Context docs](https://line-21.com/docs/ai-services/ai-context)) ## Your data Client content processed through Line 21 is not used to train any third-party AI model. | Question | Answer | | --- | --- | | Used to train AI models? | No, for any vendor in the pipeline. | | Exclude specific vendors? | Yes, per client engagement. | | EU data residency? | Yes, depending on the speech and translation providers chosen. | | Zero data retention? | Available for the language model providers. | | Recordings kept? | 30 days by default, shorter on request. | [Data handling and retention](https://line-21.com/docs/security/data-handling) --- # Pricing Start free and pay only for what you use, or choose a plan with hours included every month. Prices in USD; visitors in the UK and EU see the same figures in GBP or EUR. Machine-readable: [pricing.json](https://line-21.com/pricing.json). ## Plans ### Pay as you go — $0 / month For individual human captioners, respeakers, Velotype writers and small agencies. - Unlimited projects - Unlimited manual captioning minutes - Unlimited text destinations: browser, overlay, Zoom, Teams and YouTube - Every other service at pay as you go rates ### Essential — $199 / month For teams running regular events with AI captions. Everything in Pay as you go, plus each month - 20 hours of ASR - 20 hours of AI Proofreader - 5 translation language-hours - 1,000 HLS viewer-hours - 5 RTMP channel-hours - Pro support: email, chat and phone AI dubbing and usage beyond the included hours at pay as you go rates. ### Pro — $599 / month For productions streaming in several languages, every week. Everything in Pay as you go, plus each month - 60 hours of ASR - 60 hours of AI Proofreader - 20 translation language-hours - 5,000 HLS viewer-hours - 30 RTMP channel-hours - 10 AI dubbing language-hours - Pro support: email, chat and phone Usage beyond the included hours at pay as you go rates. ### Enterprise — tailored For organisations that need volumes, terms and support to fit. - Yearly volumes with graduated discounts - Every service, tailored to your workflow - Custom SLA and integrations - Dedicated account manager Build a custom yearly or monthly plan: [plan builder](https://line-21.com/plans.md). ## Support | Support level | Pay as you go | Essential | Pro | Enterprise | | --- | --- | --- | --- | --- | | Basic support: Email-based support. | Included | Included | Included | Included | | Pro support: Email, chat and phone support. | — | Included | Included | Included | | Enterprise support: Dedicated account manager and a custom SLA agreed with your team. | — | — | — | Included | | Live event support: A support engineer on standby during your events. Booked separately on every plan. | From $100 / hour | From $100 / hour | From $100 / hour | From $100 / hour | ## Pay as you go rates Every service is available on every plan. Pay as you go accounts are billed at these rates; paid plans use them for usage beyond their included hours. | Service | Rate | Description | | --- | --- | --- | | Manual captioning | Included | Bring your own captioners, respeakers or Velotype writers. Their captions reach browser, overlay and API token destinations such as Zoom, Teams and YouTube at no cost, with no limits. | | Human captioners | from $120 / hour | We work with a large network of professional captioners and captioning agencies to cover a wide variety of languages and timezones. | | ASR: automatic speech recognition | $15 / hour | Use speech-to-text models to automatically create captions with up to 97% accuracy. | | AI Proofreader | $21 / hour | Safeguard your captions against typos and name errors using a context-aware AI. | | AI translations | $9 / language-hour | Translate captions automatically into 120 languages. Billed for each language you add. | | AI dubbing | $30 / language-hour | Live spoken-language interpretation with synthesised voice tracks, alongside your captions and translations. | | RTMP video encoding | $24 / channel-hour | Captions encoded into your stream and sent wherever you already broadcast. | | HLS video streaming | $0.12 / viewer-hour | Go directly to your audience with an embeddable player that shows every caption language. | | Meetings | $6 / hour | Make multi-language meetings more accessible with real-time captions and translations. | | Venues | $3 / hour | Power a full-time schedule of mixed pre-recorded and live events for accessible venues. | | Live event support: standard support hours | $100 / hour | A support engineer on standby for your live event, 09:00 - 18:00 UK time. | | Live event support: extended support hours | $200 / hour | A support engineer on standby for your live event, 06:00 - 09:00 and 18:00 - 23:00 UK time. | | Live event support: out-of-hours / weekends | $300 / hour | A support engineer on standby for your live event at night and over the weekend. | Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Build your plan Configure a yearly or monthly plan: pick yearly service hours, the services you need, and support. The page shows an instant estimate and a shareable link. ## How the estimate works - Hours per year: 50–50,000. Hours are a shared pool across services, billed at the rate of the highest-priced active service. - Graduated volume discounts (each band applies only to the hours inside it): 0–250 hours at 100%; 250–1000 hours at 85%; 1000–5000 hours at 70%; 5000–20000 hours at 55%; 20000–∞ hours at 40%. - Monthly billing adds 20% to the annual price. | Service | List price per hour | Notes | | --- | --- | --- | | Live captioning + translations | $20 | Real-time captions with AI translations into any number of languages. Always on. | | Video encoding | $24 | CEA 608/708 caption encoding, SRT / RTMP ingest and simulcast output in RTMP / SRT / HLS. | | AI dubbing (interpretation) | $30 | Live spoken-language interpretation with synthesised voice tracks. Requires Video encoding. | | AI proofreader | $21 | Context-aware AI that catches typos and name errors live. Included at no extra charge. | | Support level | Surcharge on service hours | Description | | --- | --- | --- | | Basic | 0% | Email-based support. | | Pro | 15% | Email, chat and phone support. | | Enterprise | 30% | Dedicated account manager and a custom SLA agreed with your team. | | Live event support | Price per hour | Window | | --- | --- | --- | | Standard support hours | $100 | 09:00 - 18:00 UK time | | Extended support hours | $200 | 06:00 - 09:00 and 18:00 - 23:00 UK time | | Out-of-hours / weekends | $300 | Nights and weekends | ## Share links A plan is encoded in the URL: `/plans?state=`. The JSON has `v` (1), `hours`, `services` ({ `aiDubbing`, `encoding` } booleans), `standardSupport` (`basic`, `pro`, `enterprise`), `liveSupport` (hours per year for `standardHours`, `extendedHours`, `outOfHours`), `cycle` (`annual` or `monthly`), and `currency` (`USD`, `GBP`, `EUR`). Default plan: [open in the plan builder](https://line-21.com/plans?state=eyJ2IjoxLCJob3VycyI6NTAwLCJzZXJ2aWNlcyI6eyJhaUR1YmJpbmciOmZhbHNlLCJlbmNvZGluZyI6ZmFsc2V9LCJzdGFuZGFyZFN1cHBvcnQiOiJiYXNpYyIsImxpdmVTdXBwb3J0Ijp7InN0YW5kYXJkSG91cnMiOjUwLCJleHRlbmRlZEhvdXJzIjowLCJvdXRPZkhvdXJzIjowfSwiY3ljbGUiOiJhbm51YWwiLCJjdXJyZW5jeSI6IlVTRCJ9). --- # Languages Caption, translate, and dub live content across 100+ languages. Coverage varies by engine. Totals: 120 languages; human captioning 42; speech recognition 104; AI translation 93; AI dubbing 75. Machine-readable: [languages.json](https://line-21.com/languages.json). | Language | Code | Human captioning | Speech recognition engines | Translation engines | AI dubbing | Details | | --- | --- | --- | --- | --- | --- | --- | | English | EN | Available | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/english.md) | | Afrikaans | AF | Client-provided only | AWS Transcribe, Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/afrikaans.md) | | Albanian | SQ | Possible | Gladia | Azure, OpenAI | Yes | [details](https://line-21.com/languages/albanian.md) | | Amharic | AM | Client-provided only | Gladia | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/amharic.md) | | Arabic | AR | Client-provided only | Speechmatics, Gladia, Deepgram, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/arabic.md) | | Arabic (United Arab Emirates) | AR_AE | Client-provided only | Deepgram | — | No | [details](https://line-21.com/languages/arabic-united-arab-emirates.md) | | Assamese | AS | Client-provided only | Gladia, Microsoft MAI-Transcribe | — | No | [details](https://line-21.com/languages/assamese.md) | | Armenian | HY | Possible | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/armenian.md) | | Azerbaijani | AZ | Client-provided only | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/azerbaijani.md) | | Basque | EU | Available | Speechmatics, Gladia, AWS Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/basque.md) | | Bashkir | BA | Possible | Speechmatics, Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/bashkir.md) | | Belarusian | BE | Client-provided only | Speechmatics, Gladia, Deepgram | — | Yes | [details](https://line-21.com/languages/belarusian.md) | | Bengali | BN | Possible | Gladia, Deepgram, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/bengali.md) | | Tibetan | BO | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/tibetan.md) | | Breton | BR | Client-provided only | — | — | No | [details](https://line-21.com/languages/breton.md) | | Bosnian | BS | Client-provided only | Gladia, Deepgram, Microsoft MAI-Transcribe | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/bosnian.md) | | Bulgarian | BG | Client-provided only | Speechmatics, Deepgram, Gladia, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/bulgarian.md) | | Catalan | CA | Available | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/catalan.md) | | Cebuano | CEB | Possible | — | — | No | [details](https://line-21.com/languages/cebuano.md) | | Chinese (Simplified) | ZH | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/chinese-simplified.md) | | Corsican | CO | Possible | — | — | No | [details](https://line-21.com/languages/corsican.md) | | Croatian | HR | Client-provided only | Speechmatics, Gladia, Deepgram, AWS Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/croatian.md) | | Czech | CS | Client-provided only | Speechmatics, Gladia, Deepgram, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/czech.md) | | Danish | DA | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/danish.md) | | Dutch | NL | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/dutch.md) | | Esperanto | EO | Client-provided only | Speechmatics | — | No | [details](https://line-21.com/languages/esperanto.md) | | Estonian | ET | Possible | Speechmatics, Deepgram, Gladia, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/estonian.md) | | Finnish | FI | Client-provided only | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/finnish.md) | | French | FR | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/french.md) | | Faroese | FO | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/faroese.md) | | Frisian | FY | Client-provided only | — | — | No | [details](https://line-21.com/languages/frisian.md) | | Galician | GL | Client-provided only | Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/galician.md) | | Georgian | KA | Client-provided only | Gladia | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/georgian.md) | | German | DE | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/german.md) | | Greek | EL | Client-provided only | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/greek.md) | | Gujarati | GU | Client-provided only | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/gujarati.md) | | Haitian Creole | HT | Possible | Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/haitian-creole.md) | | Hausa | HA | Client-provided only | Gladia | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/hausa.md) | | Hawaiian | HAW | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/hawaiian.md) | | Hebrew | HE | Possible | Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/hebrew.md) | | Hindi | HI | Possible | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/hindi.md) | | Hmong | HMN | Client-provided only | — | — | No | [details](https://line-21.com/languages/hmong.md) | | Hungarian | HU | Possible | Speechmatics, Gladia, Deepgram, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/hungarian.md) | | Icelandic | IS | Possible | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/icelandic.md) | | Igbo | IG | Client-provided only | — | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/igbo.md) | | Indonesian | ID | Client-provided only | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/indonesian.md) | | Irish | GA | Client-provided only | — | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/irish.md) | | Italian | IT | Available | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/italian.md) | | Japanese | JA | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/japanese.md) | | Javanese | JV | Client-provided only | Gladia | — | Yes | [details](https://line-21.com/languages/javanese.md) | | Kannada | KN | Client-provided only | Gladia, Deepgram, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/kannada.md) | | Kazakh | KK | Client-provided only | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/kazakh.md) | | Khmer | KM | Client-provided only | Gladia | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/khmer.md) | | Kinyarwanda | RW | Client-provided only | — | Azure, OpenAI | Yes | [details](https://line-21.com/languages/kinyarwanda.md) | | Korean | KO | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/korean.md) | | Kurdish | KU | Client-provided only | — | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/kurdish.md) | | Kyrgyz | KY | Client-provided only | — | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/kyrgyz.md) | | Lingala | LN | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/lingala.md) | | Lao | LO | Client-provided only | Gladia | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/lao.md) | | Latin | LA | Client-provided only | Gladia | Google, OpenAI | No | [details](https://line-21.com/languages/latin.md) | | Latvian | LV | Possible | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/latvian.md) | | Lithuanian | LT | Possible | Speechmatics, Deepgram, Gladia, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/lithuanian.md) | | Luxembourgish | LB | Possible | Gladia | — | No | [details](https://line-21.com/languages/luxembourgish.md) | | Macedonian | MK | Client-provided only | Gladia, Deepgram, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/macedonian.md) | | Malagasy | MG | Client-provided only | Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/malagasy.md) | | Malay | MS | Client-provided only | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/malay.md) | | Malayalam | ML | Client-provided only | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/malayalam.md) | | Maltese | MT | Possible | Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/maltese.md) | | Maori | MI | Client-provided only | Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/maori.md) | | Marathi | MR | Client-provided only | Speechmatics, Gladia, Deepgram, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/marathi.md) | | Mongolian | MN | Client-provided only | Speechmatics, Gladia | — | Yes | [details](https://line-21.com/languages/mongolian.md) | | Myanmar (Burmese) | MY | Client-provided only | Gladia | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/myanmar-burmese.md) | | Nepali | NE | Client-provided only | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/nepali.md) | | Nynorsk | NN | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/nynorsk.md) | | Norwegian | NO | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, OpenAI | Yes | [details](https://line-21.com/languages/norwegian.md) | | Nyanja (Chichewa) | NY | Client-provided only | — | — | No | [details](https://line-21.com/languages/nyanja-chichewa.md) | | Occitan | OC | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/occitan.md) | | Odia (Oriya) | OR | Client-provided only | — | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/odia-oriya.md) | | Pashto | PS | Client-provided only | Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/pashto.md) | | Persian | FA | Client-provided only | AWS Transcribe, Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/persian.md) | | Polish | PL | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/polish.md) | | Portuguese | PT | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/portuguese.md) | | Punjabi | PA | Client-provided only | Gladia, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/punjabi.md) | | Romanian | RO | Possible | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/romanian.md) | | Russian | RU | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/russian.md) | | Samoan | SM | Client-provided only | — | Azure, OpenAI | No | [details](https://line-21.com/languages/samoan.md) | | Scots Gaelic | GD | Possible | — | — | No | [details](https://line-21.com/languages/scots-gaelic.md) | | Sanskrit | SA | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/sanskrit.md) | | Serbian | SR | Possible | AWS Transcribe, Gladia, Deepgram | — | Yes | [details](https://line-21.com/languages/serbian.md) | | Sesotho | ST | Client-provided only | — | Azure, OpenAI | No | [details](https://line-21.com/languages/sesotho.md) | | Shona | SN | Client-provided only | Gladia | Azure, OpenAI | No | [details](https://line-21.com/languages/shona.md) | | Sindhi | SD | Client-provided only | Gladia | Azure, OpenAI | Yes | [details](https://line-21.com/languages/sindhi.md) | | Sinhala (Sinhalese) | SI | Client-provided only | Gladia | Azure, OpenAI | Yes | [details](https://line-21.com/languages/sinhala-sinhalese.md) | | Slovak | SK | Possible | Speechmatics, Deepgram, Gladia, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/slovak.md) | | Slovenian | SL | Possible | Speechmatics, Gladia, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/slovenian.md) | | Somali | SO | Client-provided only | Gladia | Azure, OpenAI | No | [details](https://line-21.com/languages/somali.md) | | Spanish | ES | Available | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/spanish.md) | | Sundanese | SU | Client-provided only | Gladia | — | Yes | [details](https://line-21.com/languages/sundanese.md) | | Swahili | SW | Client-provided only | Speechmatics, Gladia, Microsoft MAI-Transcribe | Azure, OpenAI | Yes | [details](https://line-21.com/languages/swahili.md) | | Swedish | SV | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/swedish.md) | | Tagalog (Filipino) | TL | Client-provided only | AWS Transcribe, Gladia, Deepgram, Microsoft MAI-Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/tagalog-filipino.md) | | Tajik | TG | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/tajik.md) | | Tamil | TA | Client-provided only | Deepgram, Speechmatics, Gladia, Microsoft MAI-Transcribe | Azure, OpenAI | Yes | [details](https://line-21.com/languages/tamil.md) | | Tamasheq | TAQ | Client-provided only | Deepgram | — | No | [details](https://line-21.com/languages/tamasheq.md) | | Tatar | TT | Client-provided only | Gladia | Azure, OpenAI | No | [details](https://line-21.com/languages/tatar.md) | | Telugu | TE | Client-provided only | Gladia, Deepgram, Microsoft MAI-Transcribe | Azure, OpenAI | Yes | [details](https://line-21.com/languages/telugu.md) | | Thai | TH | Available | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Azure, OpenAI | Yes | [details](https://line-21.com/languages/thai.md) | | Turkish | TR | Available | Deepgram, Speechmatics, Gladia, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/turkish.md) | | Turkmen | TK | Client-provided only | Gladia | Azure, OpenAI | No | [details](https://line-21.com/languages/turkmen.md) | | Ukrainian | UK | Available | Deepgram, Speechmatics, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | DeepL, Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/ukrainian.md) | | Urdu | UR | Client-provided only | Gladia, Microsoft MAI-Transcribe | Azure, OpenAI | Yes | [details](https://line-21.com/languages/urdu.md) | | Uyghur | UG | Client-provided only | Speechmatics | Azure, OpenAI | No | [details](https://line-21.com/languages/uyghur.md) | | Uzbek | UZ | Client-provided only | Gladia | Azure, OpenAI | Yes | [details](https://line-21.com/languages/uzbek.md) | | Vietnamese | VI | Available | Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe | Azure, OpenAI | Yes | [details](https://line-21.com/languages/vietnamese.md) | | Wolof | WO | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/wolof.md) | | Welsh | CY | Client-provided only | Speechmatics, Gladia | Google, Azure, OpenAI | No | [details](https://line-21.com/languages/welsh.md) | | Xhosa | XH | Client-provided only | — | Azure, OpenAI | No | [details](https://line-21.com/languages/xhosa.md) | | Yiddish | YI | Client-provided only | Gladia | — | No | [details](https://line-21.com/languages/yiddish.md) | | Yoruba | YO | Client-provided only | Gladia | Azure, OpenAI | No | [details](https://line-21.com/languages/yoruba.md) | | Zulu | ZU | Client-provided only | AWS Transcribe | Google, Azure, OpenAI | Yes | [details](https://line-21.com/languages/zulu.md) | --- # Features Everything the Line 21 platform does for a live caption: hear the speech, check every line, carry it into other languages, and deliver it wherever your audience is. ## Features ### Speech Recognition Real-time speech-to-text from several leading engines in over 100 languages, with the best engine recommended for each language. [Learn more](https://line-21.com/features/speech-recognition) ### AI Proofreader Checks every caption line against your event's context, fixing names, brands and jargon before your audience sees it. [Learn more](https://line-21.com/features/ai-proofreader) ### Translations Live machine translation of every caption line into many languages at once, with engines and dictionaries per language. [Learn more](https://line-21.com/features/translations) ### AI Dubbing Speech-to-speech interpretation that carries the speaker's voice into another language's audio, for audiences who would rather listen than read. [Learn more](https://line-21.com/features/ai-dubbing) ### Caption Encoding Encode CEA-608 captions into your RTMP or SRT stream and simulcast it to YouTube, Twitch, Facebook, Vimeo and more. [Learn more](https://line-21.com/features/caption-encoding) ### HLS Streaming Publish your video with a caption track for every language to an HLS player you can embed on your own site. [Learn more](https://line-21.com/features/hls-streaming) ### Integrations Show captions on a browser page, an on-screen overlay, Zoom, Microsoft Teams, YouTube and other platforms. [Learn more](https://line-21.com/features/integrations) ### Monitor See your inputs, AI services and every destination at a glance while the event is live. [Learn more](https://line-21.com/features/monitor) ### Zoom App Add automated captions and machine translations to Zoom meetings and webinars with the Line 21 app for Zoom. [Learn more](https://line-21.com/zoom) ## Need people as well? The platform works on its own or alongside our team. These are the services our people provide. ### Human Live Captions Accredited stenographers and respeakers for events where every word matters. [Learn more](https://line-21.com/services/human-live-captions) ### Support Help from setup to live event support, so nothing is left to chance on the day. [Learn more](https://line-21.com/services/support) ### Post Production Subtitles and offline captions for recorded films, courses and training videos. [Learn more](https://line-21.com/services/post-production) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # AI Proofreader An AI safeguard that proofreads every caption line in real time, using the AI Context you give it about your event — so names, brands, and jargon come out right before the audience sees them. It works with human captioners and ASR alike. ## How it works ### 1. Give it context Tell the AI Context Builder about your event: chat with it, paste a script, add a link, or upload documents. Update it at any time; the proofreader uses the change from the very next line. [AI Context](https://line-21.com/docs/ai-services/ai-context) ### 2. It proofreads every line An LLM checks each caption line from your captioner or ASR engine against that context, typically in about half a second. [AI Proofreader](https://line-21.com/docs/ai-services/ai-proofreader) ### 3. Fixed before it goes out With “Required for distribution” on, captions reach your audience only once they are proofread. We recommend it for every event, and it matters most for livestreams and translations, where text can't be corrected once it's sent. [Before or after distribution](https://line-21.com/docs/ai-services/ai-proofreader#before-or-after) ## AI Context AI works best with the same preparation a human captioner would get. A single line naming the subject already helps; scripts, slide decks, agendas, and speaker lists help more. The AI Context Builder reads what you give it and keeps three things: | Type | What it holds | Example | | --- | --- | --- | | Description | What the event is about, as short or long as you like. | A conference on DNA sequencing and its uses in medicine. | | Names | Speakers, companies, products, and places that must be spelled right. | Dr. Sara Enstein · Dr. Cisco Frank | | Specialised language | Jargon and terminology specific to the subject. | nanopore · error-corrected assembly | The same context also gives ASR engines a dictionary of the words to expect. ## AI Proofreader It works best where the language is specialised or where names and brands must be spelled exactly right. What it checks: - Spelling of names, brands, places, and products - Grammar, punctuation, and formatting - Filler words and gaps in conversation - Your own rules, written in plain language What it doesn't do: - Listen to the audio — it works on the text - Replace the ASR dictionary — it runs after the engine - Check facts - Latency: <1s - Works with: Human & ASR - Engines: OpenAI & Anthropic ## Precision meets efficiency The AI Proofreader guards against goofs — accurate and efficient — polishing every line of text so your audience only laughs at captions that are meant to have a punchline. --- # Speech Recognition Real-time speech-to-text for live events from several leading engines, with the best one recommended for each language and a dictionary built from your event. - **Languages**: 104 - **Engines**: 5 - **Latency**: 3s Engines: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft MAI-Transcribe. ## What it does ### The right engine for each language Line 21 recommends the best-performing engine for every language, and you can switch engines at any time. ### Your words, recognised AI Context turns your scripts, slides and speaker names into a dictionary of words the engine should expect. [AI Context](https://line-21.com/docs/ai-services/ai-context) ### Speakers and blended languages Some engines tell speakers apart and follow conversations that switch between languages mid-sentence. [AWS Transcribe](https://line-21.com/features/speech-recognition/aws-transcribe) ### Audio from anywhere Feed audio from a browser microphone, the Line 21 desktop app, or an RTMP or SRT video stream. [Audio feeds](https://line-21.com/docs/feeds/audio) ## Learn more - [Speech recognition docs](https://line-21.com/docs/methods/asr) - [Video feeds](https://line-21.com/docs/feeds/video) - [Supported languages](https://line-21.com/languages) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # AWS Transcribe Amazon Transcribe as a speech recognition engine inside Line 21, including for conversations that switch between languages. - **Languages**: 40 - **Status**: Beta ## What it does ### Wide language support AWS Transcribe covers 40 languages, including English, Spanish, French, German, Italian, Portuguese, Japanese, Korean and Chinese. ### Blended conversations It can follow speakers who switch between languages, like Spanglish or Swenglish, and caption the dialogue as it is spoken. ### Open beta Available to everyone as a beta engine. Your usage helps us tune it for the kinds of events Line 21 captions. ## Learn more - [All speech recognition engines](https://line-21.com/features/speech-recognition) - [Speech recognition docs](https://line-21.com/docs/methods/asr) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Live Translations Every caption line translated live into as many languages as your audience needs, with the engine chosen for each language. - **Languages**: 93 - **Engines**: 4 Engines: DeepL, Google, Azure, OpenAI. ## What it does ### Many languages at once Translate one source into every language your audience reads. Viewers choose theirs on the captions page. ### The right engine per language Each language uses the engine that handles it best, and you can switch engines per language. ### Dictionaries for your terms Custom dictionaries make sure brand names, proper nouns and jargon are translated the way you want. ### Corrected before it's translated Run the AI Proofreader before captions are distributed, so its fixes carry into every language. [AI Proofreader](https://line-21.com/features/ai-proofreader) ## Learn more - [AI translation docs](https://line-21.com/docs/ai-services/ai-translations) - [Supported languages](https://line-21.com/languages) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # AI Dubbing Your speakers, heard in other languages as they talk: live speech-to-speech dubbing, in step with the video, in the player and in your streams. - **Languages**: 75 - **Outputs**: HLS / RTMP / SRT - **Status**: Beta Engines: Google Gemini Live Translate. ## What it does ### Speech to speech No captions in between: Gemini Live Translate turns the speaker's voice straight into speech in another language, translating while they are still talking. ### Follows language changes When a speaker switches language in the middle of a live stream, the dubbing follows without a restart. ### Matches the speaker The dubbed voice follows the original speaker, including their gender. ### In sync with the picture Every dubbed track stays aligned with the original video, so lips and slides line up with what viewers hear. ### Wherever your audience listens An alternative audio track per language in the HLS player, or the dubbed audio in place of the original in RTMP and SRT streams. Listening on the captions page is coming soon. [HLS streaming](https://line-21.com/features/hls-streaming) ### In beta AI dubbing is in beta and switched on per organisation. Talk to us to enable it for your events. [Talk to us](https://line-21.com/contact) ## Learn more - [AI dubbing docs](https://line-21.com/docs/ai-services/ai-dubbing) - [Languages with AI dubbing](https://line-21.com/languages) - [How we use AI](https://line-21.com/ai) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Caption Encoding Closed captions encoded into your live video and simulcast to the platforms your audience watches, from YouTube to Twitch. - **Standard**: CEA-608 - **Input**: RTMP / SRT - **Output**: RTMP / SRT / HLS ## What it does ### Captions inside the stream Your video is re-encoded with a caption track that the platform shows as closed captions, ready for viewers to switch on. ### Synchronised captions The stream is held back about 10 seconds so each caption lands on the picture at the moment its words were spoken, instead of trailing behind the speaker as live captions usually do. ### Simulcast to every platform YouTube, Facebook, X, Twitch, Vimeo, Dailymotion, Kaltura, Brightcove, Microsoft Teams, Wowza, Akamai and custom RTMP endpoints. ### Encoded in the cloud No encoder hardware to rent or run: Line 21 handles the encoding as a cloud service. ### Good to know Platforms accept one caption track, the stream reaches viewers about 10 seconds later than the live feed, and Latin-based scripts are the most reliable. For several languages, use HLS streaming. [HLS streaming](https://line-21.com/features/hls-streaming) ## Learn more - [RTMP streaming docs](https://line-21.com/docs/destinations/rtmp-streaming) - [Video feeds](https://line-21.com/docs/feeds/video) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # HLS Streaming Publish your live video with a caption track for every language, and dubbed audio tracks too, in our player or yours, delivered through a CDN to any audience size. - **Caption tracks**: All languages - **Input**: RTMP / SRT - **Replay**: 30 days ## What it does ### A caption track per language Every translation is its own native caption track, picked from the player's settings menu. ### Dubbed audio tracks Add AI dubbing and each language also gets its own audio track, so viewers can listen in their language as well as read it. [AI dubbing](https://line-21.com/features/ai-dubbing) ### Our player or yours Embed the Line 21 player on your site, share a direct viewing page, or play the stream in your own player. ### Built to scale Streams are cached on a global CDN, so audience size isn't something you have to plan around. ### Captions and audio tracks in sync Caption and dubbed audio tracks stay in step with the picture, and you can delay or advance the captions to fine-tune it. ### Replay after the event Recorded streams stay available on demand for up to 30 days. ## Learn more - [HLS streaming docs](https://line-21.com/docs/destinations/hls-streaming) - [Video feeds](https://line-21.com/docs/feeds/video) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Integrations Show live captions wherever your audience is: a captions page, an on-screen overlay, inside Zoom and Teams, or in the video stream itself. - **Destination types**: 5 - **Captions page**: Included - **RTMP platforms**: 15+ ## What it does ### Captions page Every project has a shareable page at no extra cost. Viewers choose their language; you set the colours, fonts and size. Show it on a venue screen or share a QR code. [Browser page](https://line-21.com/docs/destinations/browser) ### Overlay A floating, borderless window from the Line 21 app that puts captions on top of slides or any other application. [Overlay](https://line-21.com/docs/line-21-app/overlay) ### Zoom, Teams and YouTube Send captions as text into Zoom and Microsoft Teams with the meeting's caption token, or to YouTube Live through its caption ingestion URL. [Text destinations](https://line-21.com/docs/destinations/text-destinations) ### Encoded into the stream Closed captions inside your RTMP stream, simulcast to YouTube, Twitch, Facebook and other platforms. [Caption encoding](https://line-21.com/features/caption-encoding) ### HLS player Your video with a caption track for every language, in our player or yours. [HLS streaming](https://line-21.com/features/hls-streaming) ## Learn more - [Destinations docs](https://line-21.com/docs/destinations) - [Line 21 app for Zoom](https://line-21.com/zoom) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Monitor One screen for your whole captioning workflow while the event is live: what's connected, what's working, and where attention is needed. ## What it does ### Inputs See every active speech recognition connection and its source, and check the health of your video stream. ### Captioners Check that a captioner is connected and writing, and chat with the Line 21 support team inside the project. ### AI services Follow captions and translations as they are produced, and watch the AI Proofreader's corrections as they happen. ### Destinations Make sure captions are reaching every destination your audience is watching. ## Learn more - [Monitor docs](https://line-21.com/docs/project-management/monitor) - [Live event support](https://line-21.com/services/support) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Solutions Line 21 by the kind of event you run: the people, AI and delivery options that suit it, from a single keynote to a festival weekend. ## Conferences and keynotes Captions on the big screen and in every stream, from a single keynote to a multi-track, multilingual conference. Covers: Keynotes and speakers, Panels and discussions, International conferences, Summits and expos, Awards ceremonies. [More about Conferences and keynotes](https://line-21.com/solutions/conferences-and-keynotes) ## Festivals and sports Big screens, loud venues and long days: captions for announcements, performances and live commentary. Covers: Festivals, Cultural events, Sports events, Venues. [More about Festivals and sports](https://line-21.com/solutions/festivals-and-sports) ## Corporate communications Town halls, all-hands and internal broadcasts that every employee can follow, in their own language. Covers: Town halls, All-hands meetings, Corporate meetings, Corporate calls, Internal broadcasts. [More about Corporate communications](https://line-21.com/solutions/corporate-communications) ## Online meetings and webinars Captions and translations inside Zoom and Teams, or on a shared page next to any other platform. Covers: Zoom and Microsoft Teams calls, Webinars, Virtual events, Multilingual meetings. [More about Online meetings and webinars](https://line-21.com/solutions/online-meetings-and-webinars) ## Education Lectures, seminars and online courses that every student can follow, in the room or remotely. Covers: Lectures and classrooms, Seminars, Online courses. [More about Education](https://line-21.com/solutions/education) ## Streaming and broadcast Closed captions in your live stream, on every platform you simulcast to or in your own HLS player. Covers: YouTube Live, Twitch, Vimeo, LinkedIn Live, Facebook Live, Your own player. [More about Streaming and broadcast](https://line-21.com/solutions/streaming-and-broadcast) [Custom solutions](https://line-21.com/services/custom) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Conferences and keynotes Conference audiences follow along in the room, on their phones and in the livestream at the same time. Line 21 captions each session once and sends it to all of them, in as many languages as your audience speaks. Book professional captioners for the sessions that matter most, use AI for the breakout rooms, or mix the two across the programme. Covers: Keynotes and speakers, Panels and discussions, International conferences, Summits and expos, Awards ceremonies. ## What makes it hard **Names, brands and jargon**: Speakers rattle off product names, acronyms and the names of fellow panellists. Generic captions get them wrong at the worst moment. **Several audiences at once**: The room needs captions on screen, remote viewers need them in the stream, and international guests need their own language. **Fast exchanges**: Panels, Q&A and award announcements move quickly, between several speakers, with no second take. ## How Line 21 handles it ### Brief the AI like a captioner Add the agenda, slides and speaker list. The AI Proofreader uses them to fix names and terms in every line before it is shown. [AI Proofreader](https://line-21.com/features/ai-proofreader) ### Professional captioners where it counts Accredited stenographers and respeakers for keynotes and high-stakes sessions, working on the same platform as the AI. [Human live captions](https://line-21.com/services/human-live-captions) ### Every language in the room Each caption line is translated live, and attendees pick their language on the captions page. [Translations](https://line-21.com/features/translations) ### Screens, phones and streams Show captions on venue screens with the browser page or the overlay, share a QR code for phones, and add them to your livestream. [Integrations](https://line-21.com/features/integrations) ## Further reading - [5 tools for live captioning at keynotes and speaker sessions](https://line-21.com/blog/2026-07-03-5-tools-for-live-captioning-at-keynotes) - [How to provide in-room captions at live events](https://line-21.com/blog/2026-07-24-how-to-provide-in-room-captions-at-live-events) - [6 live translation providers for conferences](https://line-21.com/blog/2026-08-29-6-live-translation-providers-for-conferences) - [5 tools for live captioning at awards ceremonies](https://line-21.com/blog/2026-08-17-5-tools-for-live-captioning-at-awards-ceremonies) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Festivals and sports Large venues bring their own captioning problems: crowd noise, long days across several stages, and audiences who follow on the big screen or on their own phones. Line 21 sends captions to venue screens, to the stream and to a link anyone can open, and can mix live captions with pre-recorded content on a schedule. Covers: Festivals, Cultural events, Sports events, Venues. ## What makes it hard **Noisy audio**: Captions are only as good as the audio they hear. A clean feed from the mixing desk beats a microphone in the crowd. **Long, scheduled days**: Stages run from morning to night, with live sessions between pre-recorded spots and announcements. **Audiences on the move**: People move between stages and want captions on their own phones, not only on a fixed screen. ## How Line 21 handles it ### Straight from the desk Send audio from the mixer, or your RTMP or SRT programme feed, and speech recognition or a captioner takes it from there. [Audio and video feeds](https://line-21.com/docs/feeds/audio) ### Live and pre-recorded, on a schedule Playlists mix live captions with .srt or .vtt files. Viewers scan a QR code to see what is live now and what comes next. [Playlists](https://line-21.com/docs/project-types/playlist) ### Big screens and phones Every project has a captions page at no extra cost, styled to match your event, ready for a venue screen or a QR code. [Browser page](https://line-21.com/docs/destinations/browser) ### Fans watching from home Encode captions into the broadcast or stream so remote viewers get them too. [Caption encoding](https://line-21.com/features/caption-encoding) ## Further reading - [5 tools for live captioning at festivals and large-scale events](https://line-21.com/blog/2026-07-18-5-tools-for-live-captioning-at-festivals) - [Meet the moment: live captioning display choices](https://line-21.com/blog/meet-the-moment) - [How to provide in-room captions at live events](https://line-21.com/blog/2026-07-24-how-to-provide-in-room-captions-at-live-events) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Corporate communications Company-wide meetings only work if everyone can follow them, including colleagues who are deaf or hard of hearing and those working in a second language. Line 21 captions the meeting once and delivers it in every language your teams speak, on the platforms you already use. Covers: Town halls, All-hands meetings, Corporate meetings, Corporate calls, Internal broadcasts. ## What makes it hard **Global teams**: One town hall, many languages. Employees follow more easily, and take more in, in their own. **Confidential content**: Internal announcements need captioning without your content ending up as someone's training data. **The tools you already use**: Meetings run on Teams, Zoom or an internal streaming platform, and captions need to appear there. ## How Line 21 handles it ### Every team's language Live translation into many languages at once. Each employee chooses theirs on the captions page. [Translations](https://line-21.com/features/translations) ### Inside Teams and Zoom Send captions straight into Microsoft Teams and Zoom with the meeting's caption token, or encode them into your internal stream. [Integrations](https://line-21.com/features/integrations) ### Captioners for sensitive sessions For leadership announcements, accredited captioners who are bound by professional confidentiality. [Human live captions](https://line-21.com/services/human-live-captions) ### Your content stays yours Content that goes through Line 21 is not used to train third-party AI models, and vendors can be restricted where your policy requires it. [Data handling](https://line-21.com/docs/security/data-handling) ## Further reading - [5 tools for live captioning at corporate events](https://line-21.com/blog/2026-08-12-5-tools-for-live-captioning-at-corporate-events) - [5 best AI translation tools for corporate town halls](https://line-21.com/blog/2026-07-16-5-best-ai-translation-tools-for-corporate-town-halls) - [Structuring for scale: captioning for multilingual events](https://line-21.com/blog/structuring-for-scale) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Online meetings and webinars Online events reach people wherever they are, but only if they can follow along. Line 21 adds live captions and translations to the meeting tools your audience already uses. Send captions into Zoom or Microsoft Teams, let an AI call agent join and transcribe, or share a captions page that works next to any platform. Covers: Zoom and Microsoft Teams calls, Webinars, Virtual events, Multilingual meetings. ## What makes it hard **Captions inside the call**: Attendees expect captions in the meeting window, not in a separate tab they have to find. **More than one spoken language**: Multilingual calls have several source languages, and every participant needs to follow all of them. **No AV team**: Webinars are often run by one person who needs a setup that simply works. ## How Line 21 handles it ### Captions in Zoom and Teams Send live captions into Zoom and Microsoft Teams with the meeting's caption token, so they appear in the meeting window. [Integrations](https://line-21.com/features/integrations) ### An AI call agent A bot joins your Zoom call and transcribes it live, translating as it goes. [AI call agent](https://line-21.com/docs/ai-services/ai-call-agent) ### Meetings in several languages Meeting projects let each participant speak their own language and read everyone else in theirs. [Meeting projects](https://line-21.com/docs/project-types/meeting) ### The Line 21 app for Zoom Add automated captions and machine translations to Zoom meetings and webinars from the Zoom App Marketplace. [Zoom app](https://line-21.com/zoom) ## Further reading - [How to live caption into Zoom via API token](https://line-21.com/blog/2026-07-07-live-caption-into-zoom-via-api-token) - [5 tools for live translation on Zoom webinars](https://line-21.com/blog/2026-08-14-5-tools-for-live-translation-on-zoom-webinars) - [5 tools for live captioning virtual events](https://line-21.com/blog/2026-08-13-5-tools-for-live-captioning-virtual-events) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Education Students who are deaf or hard of hearing, studying in a second language, or sitting at the back of a large hall all learn better with captions. Line 21 brings professional captioners or speech recognition into the lecture hall and the online classroom, with translations for international students. Covers: Lectures and classrooms, Seminars, Online courses. ## What makes it hard **Specialist vocabulary**: Lectures are full of technical terms that general-purpose speech recognition has never heard. **Accessibility obligations**: Captioning is often a reasonable adjustment under accessibility law, where human captioners are the recognised standard. **Recorded content too**: Lectures are recorded and reused, and the recordings need subtitles as well. ## How Line 21 handles it ### Recognised human captioning CART captioners recognised under regulations such as the ADA, the Equality Act 2010 and the European Accessibility Act. [Human live captions](https://line-21.com/services/human-live-captions) ### AI that knows the subject Add the syllabus, slides and reading lists so speech recognition and the AI Proofreader know the course's terms. [AI Proofreader](https://line-21.com/features/ai-proofreader) ### On every student's device Students follow on their own laptop or phone and choose the language they read in. [Browser page](https://line-21.com/docs/destinations/browser) ### Subtitles for recordings Accurate, well-timed subtitles for recorded lectures and course videos. [Post production](https://line-21.com/services/post-production) ## Further reading - [Keeping human captioners in control](https://line-21.com/blog/2026-05-28-keeping-human-captioners-in-control) - [7 human CART providers and platforms compared](https://line-21.com/blog/2026-08-03-7-human-cart-providers-and-platforms-compared) - [Accessible live captions: talent and technology](https://line-21.com/blog/talent-and-technology) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Streaming and broadcast Every streaming platform has its own rules for captions. Line 21 handles them from one place: send your stream in, and get it back with captions wherever your audience watches. Encode captions for social platforms, post them to YouTube through its caption API, or publish your own player with a caption track for every language. Covers: YouTube Live, Twitch, Vimeo, LinkedIn Live, Facebook Live, Your own player. ## What makes it hard **One caption track**: Most social platforms accept a single caption track in the stream, so one language at a time. **Platform by platform**: YouTube, Twitch, Facebook and Vimeo each set up captions differently. **Delay and sync**: Encoding adds delay, and captions have to stay in step with the picture. ## How Line 21 handles it ### Encoded and simulcast CEA-608 captions encoded into your RTMP or SRT feed and sent to YouTube, Twitch, Facebook, X, Vimeo or any custom RTMP endpoint. [Caption encoding](https://line-21.com/features/caption-encoding) ### Your own player, every language An embeddable HLS player with a caption track per language, delivered through a CDN to any number of viewers. [HLS streaming](https://line-21.com/features/hls-streaming) ### YouTube without re-encoding Send timestamped captions to YouTube Live through its caption ingestion URL. [Text destinations](https://line-21.com/docs/destinations/text-destinations) ### Eyes on every output Watch the captioned video and every destination from one screen while you are live. [Monitor](https://line-21.com/features/monitor) ## Further reading - [5 tools for live captioning on YouTube Live](https://line-21.com/blog/2026-07-06-5-tools-for-live-captioning-on-youtube-live) - [5 tools for closed captioning across multiple streaming destinations](https://line-21.com/blog/2026-07-29-5-tools-for-closed-captioning-across-multiple-streaming-destinations) - [Live captioning for HLS video players: comparing 5 platforms](https://line-21.com/blog/2026-08-03-live-captioning-for-hls-video-players-comparing-5-platforms) - [5 tools for live captioning on Twitch](https://line-21.com/blog/2026-07-10-5-tools-for-live-captioning-on-twitch) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Line 21 compared How Line 21 compares with other live captioning solutions, side by side, based on what each company publishes about itself. - [Line 21 vs Ai-Media](https://line-21.com/compare/ai-media): A long-established captioning company with its LEXI AI, iCap network and broadcast encoders, compared with Line 21's cloud platform. - [Line 21 vs Verbit](https://line-21.com/compare/verbit): An AI and human transcription company that owns VITAC, compared with Line 21's open live captioning platform. - [Line 21 vs SyncWords](https://line-21.com/compare/syncwords): A cloud service for AI captions, translated subtitles and AI dubbing in live streams, compared with Line 21. - [Line 21 vs CaptionHub](https://line-21.com/compare/captionhub): An enterprise subtitling and voiceover platform with a live subtitles product, compared with Line 21's live-first platform. - [Line 21 vs StreamText](https://line-21.com/compare/streamtext): A long-standing caption delivery service for CART captioners, compared with Line 21's captioning platform. --- # Line 21 vs Ai-Media Ai-Media and Line 21 both caption live broadcasts, streams and events with AI and human captioners. Ai-Media brings two decades of broadcast captioning and its own encoder hardware; Line 21 runs the whole workflow as one cloud platform, with a choice of AI engines and an AI Proofreader. ## At a glance | | Line 21 | Ai-Media | | --- | --- | --- | | Live AI captions | Yes, in 100+ languages | Yes: LEXI | | Human (CART) captioners | Yes: accredited stenographers and respeakers, or your own captioners, free | Yes: Premium Captions by Ai-Media captioners | | Speech recognition engines | Choose per language: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft | Its own engine, LEXI | | Correcting AI captions | AI Proofreader checks every line against your event's names and terms | Not stated publicly | | Live translation | Yes, 90+ languages with DeepL, Google, Azure and OpenAI | Yes: iCap Translate and LEXI | | Live AI dubbing | Yes (beta): 75 languages with Gemini Live Translate | Yes: LEXI Voice | | Captions encoded into streams | Yes: CEA-608 into RTMP and SRT, in the cloud | Yes: Falcon cloud encoder and broadcast encoder hardware | | Multi-language HLS player | Yes: a caption and dubbed audio track per language | Not stated publicly | | Zoom and Teams | Yes: caption token for Zoom and Teams, plus a Zoom app | Not stated publicly | | Pricing | Free pay as you go; plans from $199 a month | Subscriptions for Falcon and encoders | ## Key differences **One platform or a set of products**: Ai-Media's offer spans LEXI, iCap, Falcon and hardware encoders. Line 21 runs captioning, translation, dubbing, encoding and HLS delivery in one cloud project, with no hardware to install. **Choice of AI**: Ai-Media uses its own LEXI engine. Line 21 recommends the best of several speech engines for each language and lets you switch. **Correcting AI captions live**: Line 21's AI Proofreader checks each caption line against your agenda, speakers and terminology before it is shown, for human and AI captions alike. **Broadcast heritage**: Ai-Media has deep roots in broadcast and its own encoder hardware, which can matter for traditional broadcast plants built around SDI encoders. ## Choose Line 21 if - You want captions, translation, dubbing and streaming in one cloud platform - You want to pick the speech engine per language - Names and jargon must be right, with live AI proofreading - You bring your own captioners, or want a free platform for them ## Choose Ai-Media if - Your broadcast plant is built around hardware caption encoders - You already use the iCap network across your channels ## About Ai-Media Ai-Media has worked in captioning and accessibility since 2003. Its LEXI engine produces AI captions, LEXI Voice translates speech into other languages, and its human captioners deliver Premium Captions. Captions travel over its iCap network, with Falcon as a cloud RTMP encoder and, from 2026, new LEXI Text and Voice encoders sold on subscription. Sources: - [Ai-Media brochure, April 2026](https://landing.ai-media.tv/hubfs/AIM%20Brochure_EN_2026APR_DIGITAL_COMPRESSED%20(1).pdf) - [Ai-Media caption delivery and encoders](https://www.ai-media.tv/our-products/caption-delivery/) - [LEXI Text and LEXI Voice encoders at NAB 2026](https://bernama.com/en/press/news.php?id=2545057) Based on Ai-Media's public information, reviewed October 2026. Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Line 21 vs Verbit Verbit and Line 21 both combine AI with human captioners for live events. Verbit runs a large managed transcription service, strong in education, legal and broadcast through VITAC; Line 21 is a platform you run yourself, with any captioners you choose and a choice of AI engines. ## At a glance | | Line 21 | Verbit | | --- | --- | --- | | Live AI captions | Yes, in 100+ languages | Yes: its own speech recognition | | Human (CART) captioners | Yes: accredited stenographers and respeakers, or your own captioners, free | Yes: Verbit's transcribers and CART captioners | | Speech recognition engines | Choose per language: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft | Its own engine | | Correcting AI captions | AI Proofreader checks every line against your event's names and terms | Human transcribers review AI output | | Live translation | Yes, 90+ languages with DeepL, Google, Azure and OpenAI | Not stated publicly for live events | | Live AI dubbing | Yes (beta): 75 languages with Gemini Live Translate | Not stated publicly | | Captions encoded into streams | Yes: CEA-608 into RTMP and SRT, in the cloud | Broadcast captioning through its VITAC business | | Multi-language HLS player | Yes: a caption and dubbed audio track per language | Not stated publicly | | Zoom and Teams | Yes: caption token for Zoom and Teams, plus a Zoom app | Yes: Zoom through a caption token | | Pricing | Free pay as you go; plans from $199 a month | Not published | ## Key differences **Managed service or open platform**: Verbit supplies its own transcribers as a service. On Line 21 you can book our captioners, bring your own, or run AI alone, all in the same project. **Choice of AI**: Verbit uses its own speech recognition. Line 21 lets you choose from several engines per language. **Translation and dubbing**: Line 21 translates every caption line live into 90+ languages and can dub speech into 75 languages, on the same platform. **Delivery**: Line 21 encodes captions into RTMP and SRT streams and publishes multi-language HLS itself, with no separate broadcast service. ## Choose Line 21 if - You want to run captioning yourself, with your own or our captioners - You need live translation or AI dubbing alongside captions - You stream to several platforms and want captions in each - You want published, predictable pricing ## Choose Verbit if - You want a fully managed transcription service across many courses or hearings - Your captioning is tied to US broadcast workflows already served by VITAC ## About Verbit Verbit combines its own speech recognition with human transcribers, and offers live CART captioning that can be delivered into Zoom. In 2021 it acquired VITAC, one of North America's largest broadcast captioning providers. Sources: - [Verbit real-time transcription (CART) announcement](https://verbit.ai/article/verbit-launches-ai-powered-real-time-transcription-solution) - [Verbit and Zoom integration](https://info.verbit.ai/hubfs/Partners%20and%20Integrations/Integrations_Verbit&Zoom_3.pdf) - [Verbit acquires VITAC (TV Technology)](https://www.tvtechnology.com/news/verbit-acquires-vitac-captioning-provider) Based on Verbit's public information, reviewed October 2026. Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Line 21 vs SyncWords SyncWords and Line 21 are the closest alternatives here: both are cloud platforms that add AI captions, translations and AI dubbing to live streams. The differences are in human captioning, speech engines, live proofreading and how you reach audiences in the room and in meetings. ## At a glance | | Line 21 | SyncWords | | --- | --- | --- | | Live AI captions | Yes, in 100+ languages | Yes, in about 50 source languages | | Human (CART) captioners | Yes: accredited stenographers and respeakers, or your own captioners, free | Accepts captions from standard CART systems | | Speech recognition engines | Choose per language: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft | Not stated publicly | | Correcting AI captions | AI Proofreader checks every line against your event's names and terms | Custom dictionaries and glossaries | | Live translation | Yes, 90+ languages with DeepL, Google, Azure and OpenAI | Yes, into about 100 languages | | Live AI dubbing | Yes (beta): 75 languages with Gemini Live Translate | Yes: 900+ AI voices in 50+ languages | | Captions encoded into streams | Yes: CEA-608 into RTMP and SRT, in the cloud | Yes: RTMP(S) with 608 captions | | Multi-language HLS player | Yes: a caption and dubbed audio track per language | Yes | | Zoom and Teams | Yes: caption token for Zoom and Teams, plus a Zoom app | Not stated publicly | | Pricing | Free pay as you go; plans from $199 a month | Not published | ## Key differences **Human captioners built in**: Line 21 supplies accredited stenographers and respeakers and connects their CART software directly, alongside AI, in the same project. **Live proofreading**: Beyond dictionaries, Line 21's AI Proofreader reads every caption line against your event's context and fixes names and terms before viewers see them. **Beyond the stream**: Line 21 also delivers to venue screens, an on-screen overlay, a shareable captions page, and Zoom and Teams meetings. **Choice of speech engines**: Line 21 recommends the best of several engines for each language and lets you switch per event. ## Choose Line 21 if - You mix human captioners and AI across your events - Your audience is in the room and in meetings, not only watching a stream - You want live AI proofreading of names and terminology - You want published pricing and a free starting tier ## Choose SyncWords if - You need a large choice of AI dub voices - Your streams run on AWS Elemental MediaLive and MediaPackage ## About SyncWords SyncWords turns live sources into same-language captions, translated subtitles and AI-voiced dubs. It takes SRT, RTMP, HLS and CMAF inputs, outputs to HLS players and RTMP with 608 captions, and integrates with AWS Elemental workflows. Human captions can arrive through standard CART systems. Sources: - [SyncWords Live: captions, subtitles and AI dubbing](https://www.syncwords.com/products/syncwords-live/) - [SyncWords captions and translations using HLS](https://syncwords.com/products/automated-captions-translations-using-hls) - [SyncWords with AWS Elemental](https://www.syncwords.com/solutions/live-captions-translations-aws-elemental) Based on SyncWords's public information, reviewed October 2026. Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Line 21 vs CaptionHub CaptionHub is built around enterprise subtitling, translation and voiceover for recorded video, with CaptionHub Live adding AI subtitles to live streams. Line 21 is built for live events first: human and AI captioning, live translation and dubbing, and delivery into the stream itself. ## At a glance | | Line 21 | CaptionHub | | --- | --- | --- | | Live AI captions | Yes, in 100+ languages | Yes: CaptionHub Live | | Human (CART) captioners | Yes: accredited stenographers and respeakers, or your own captioners, free | Not stated publicly for live events | | Speech recognition engines | Choose per language: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft | Not stated publicly | | Correcting AI captions | AI Proofreader checks every line against your event's names and terms | Termbases, dictionaries and approval workflows | | Live translation | Yes, 90+ languages with DeepL, Google, Azure and OpenAI | Yes, into 90+ languages | | Live AI dubbing | Yes (beta): 75 languages with Gemini Live Translate | Voiceover for recorded video | | Captions encoded into streams | Yes: CEA-608 into RTMP and SRT, in the cloud | No: subtitles are delivered alongside the stream | | Multi-language HLS player | Yes: a caption and dubbed audio track per language | Through supported players, such as Brightcove | | Zoom and Teams | Yes: caption token for Zoom and Teams, plus a Zoom app | Not stated publicly | | Pricing | Free pay as you go; plans from $199 a month | Not published | ## Key differences **Live first or recorded first**: CaptionHub's strength is subtitling and voiceover workflows for recorded content. Line 21 is designed for live events, from the speaker's microphone to every screen. **Captions in the stream**: Line 21 encodes captions into RTMP and SRT streams for YouTube, Twitch and other platforms, as well as publishing them alongside. **Human captioners**: Line 21 supplies accredited human captioners for live events and connects their CART software directly. **Live dubbing**: Line 21 dubs speech into other languages live, in sync with the video, rather than after the event. ## Choose Line 21 if - Your priority is live events and live streams - You need human captioners or captions encoded into the stream - You want live AI dubbing during the event ## Choose CaptionHub if - Most of your work is subtitling and voiceover for recorded video - You need approval workflows across large localisation teams ## About CaptionHub CaptionHub is an AI platform for multilingual subtitling and voiceover, with approval workflows, roles and ISO 27001 certification. CaptionHub Live adds AI captions and translations to live streams without sitting in the stream path, delivering subtitles through a CDN. CaptionHub Voiceover turns captions into synthetic dubbed audio. Sources: - [CaptionHub Live: real-time media localisation](https://www.captionhub.com/features/live) - [Introducing CaptionHub Live (MultiLingual)](https://multilingual.com/introducing-captionhub-live-synchronized-multilingual-subtitling-for-live-streams/) - [CaptionHub](https://captionhub.com/) Based on CaptionHub's public information, reviewed October 2026. Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Line 21 vs StreamText StreamText and Line 21 both give human captioners a way to deliver live text to an audience, and both now add AI captions. StreamText focuses on caption delivery with per-minute pricing; Line 21 adds a choice of AI engines, live proofreading, translation, dubbing and streaming in one platform. ## At a glance | | Line 21 | StreamText | | --- | --- | --- | | Live AI captions | Yes, in 100+ languages | Yes | | Human (CART) captioners | Yes: accredited stenographers and respeakers, or your own captioners, free | Yes: a delivery platform for CART captioners | | Speech recognition engines | Choose per language: Speechmatics, Deepgram, Gladia, AWS Transcribe, Microsoft | Not stated publicly | | Correcting AI captions | AI Proofreader checks every line against your event's names and terms | Not stated publicly | | Live translation | Yes, 90+ languages with DeepL, Google, Azure and OpenAI | Yes | | Live AI dubbing | Yes (beta): 75 languages with Gemini Live Translate | Not stated publicly | | Captions encoded into streams | Yes: CEA-608 into RTMP and SRT, in the cloud | Yes: 608/708 cloud encoding, priced separately | | Multi-language HLS player | Yes: a caption and dubbed audio track per language | Not stated publicly | | Zoom and Teams | Yes: caption token for Zoom and Teams, plus a Zoom app | Yes: Zoom, Webex and Adobe Connect | | Pricing | Free pay as you go; plans from $199 a month | Per minute, no subscription | ## Key differences **Free for captioners**: Line 21 is free for individual captioners and the agencies that book them; StreamText charges per minute of use. **AI that knows your event**: Line 21 chooses from several speech engines and checks every line with the AI Proofreader against your event's names and terms. **Translation and dubbing**: Line 21 translates captions live into 90+ languages and can dub speech into 75 languages. **Streams and players**: Line 21 includes CEA-608 encoding and a multi-language HLS player, alongside the captions page and overlay. ## Choose Line 21 if - You are a captioner looking for a free platform - You want AI captions that know your event's vocabulary - You need translation, dubbing or video streaming in the same place ## Choose StreamText if - Your clients already expect StreamText links - You need Webex or Adobe Connect integrations ## About StreamText StreamText has been used by CART captioners and court reporting firms since 2003 to deliver human-written captions. It offers an embeddable caption player, integrations with Zoom, Webex and Adobe Connect, translation, AI captioning, and a separately priced 608/708 cloud encoding service, all paid per minute. Sources: - [StreamText features and pricing](https://streamtext.net/features-and-pricing/) - [StreamText human captions](https://streamtext.net/human-captions/) - [StreamText 608/708 cloud encoding](https://streamtext.net/cloud-encoding-service/) Based on StreamText's public information, reviewed October 2026. Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Services The people behind Line 21: professional captioners, a support team that knows live events, and subtitling specialists, on their own or alongside the platform's AI. ## Features ### Human Live Captions Accredited stenographers and respeakers caption your event live, verbatim, in the room or remotely. [Learn more](https://line-21.com/services/human-live-captions) ### Support From onboarding and setup checks to a specialist on call during your live event, from $100 per hour. [Learn more](https://line-21.com/services/support) ### Post Production Accurate, well-timed subtitles and offline captions for films, TV, courses and training videos. [Learn more](https://line-21.com/services/post-production) ### Custom Solutions A setup designed around your event, platform and audience, combining delivery methods where needed. [Learn more](https://line-21.com/services/custom) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Human Live Captions Accredited stenographers and respeakers caption your event live and verbatim, in the room or remotely, for the sessions where every word matters. - **Languages**: 21 - **Speed**: 350 wpm - **Methods**: Steno / Respeaking ## What you get ### Accredited professionals CART captioners recognised under regulations such as the ADA, the Equality Act 2010 and the European Accessibility Act, and bound by professional confidentiality. ### Stenographers and respeakers Stenographers write on specialist keyboards at up to 350 words a minute; respeakers use voice software to a similar standard. Both connect to Line 21 from their own tools. [CART connector](https://line-21.com/docs/line-21-app/cart-connector) ### Prepared for your event Captioners do their best work with the agenda, speaker names and terminology in advance. Share them early and accuracy follows. ### A second pair of eyes The AI Proofreader checks human captions too, catching slips in names and terms before your audience sees them. [AI Proofreader](https://line-21.com/features/ai-proofreader) ### In every language and destination Human captions can be translated live and sent everywhere the AI's can: the captions page, overlays, Zoom, Teams and your streams. [Translations](https://line-21.com/features/translations) ### Book early Demand for captioners outstrips supply in many languages. Ask early, and plan for a pair of captioners on sessions longer than an hour. ## Learn more - [Human captioning docs](https://line-21.com/docs/methods/human) - [Languages with human captioners](https://line-21.com/languages) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Support From onboarding to a specialist on standby during your live event, so nothing is left to chance on the day. - **Live event support**: from $100/hr - **Onboarding**: Included ## What you get ### Onboarding and training Guided setup of your account and first events. Training sessions for your team. ### Basic support Email-based support. Included on every plan, pay as you go too. ### Pro support Email, chat and phone support. Included in the Essential and Pro plans. ### Enterprise support Dedicated account manager and a custom SLA agreed with your team. ### Live event support A support engineer on standby during your event, booked by the hour on any plan: $100 an hour from 09:00 - 18:00 UK time, $200 for 06:00 - 09:00 and 18:00 - 23:00 UK time, $300 at nights and weekends. [Pricing](https://line-21.com/pricing) ### Chat inside your project Every project has its own chat with the Line 21 support team, right next to your captions. [Project chat](https://line-21.com/docs/project-management/chat) ## Learn more - [Plans and support levels](https://line-21.com/pricing) - [Monitor](https://line-21.com/features/monitor) - [Documentation](https://line-21.com/docs) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Post Production Subtitles and offline captions for recorded content, from films and TV to courses and training videos: accurate, readable and well timed. - **Formats**: SRT / VTT ## What you get ### For anything recorded Films, TV, online video, lectures and corporate training: we caption and subtitle what you've already recorded. ### Accessible and compliant Captions that meet accessibility standards and serve viewers who are deaf or hard of hearing. ### Readable and well timed Sensible line breaks, comfortable reading speeds and timing that follows the speech. ### In other languages Translated subtitles adapted to each language and its audience. ### Play them live too Have a subtitle file already? Play .srt or .vtt captions during an event, or schedule them in a playlist between live sessions. [Caption playback](https://line-21.com/docs/methods/playback) ## Learn more - [Caption playback docs](https://line-21.com/docs/methods/playback) - [Playlists](https://line-21.com/docs/project-types/playlist) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Custom Solutions A setup designed around your event, platform and audience, combining delivery methods and integrations where one size doesn't fit. ## What you get ### Built around your systems We connect to the platforms and systems you already use. If you need an integration we don't have yet, ask us. [Integrations](https://line-21.com/features/integrations) ### Several delivery methods at once Encode captions for remote viewers while an overlay shows them in the room, or pair a captions page with an HLS player. ### Adapt during the event Switch delivery options as needs change, without stopping the captions. ### Live and afterwards Captions in the moment, and recordings with captions after the event. [HLS streaming](https://line-21.com/features/hls-streaming) ### Planned with our team We design, test and run the setup with you, from the first call to the live event. [Support](https://line-21.com/services/support) ## Learn more - [Destinations docs](https://line-21.com/docs/destinations) - [All features](https://line-21.com/features) Talk to us: [contact page](https://line-21.com/contact) or hello@line-21.com. --- # Blog | Date | Title | Summary | | --- | --- | --- | | 2026-10-21 | [Accessible Videos - More Options, or More Obstacles?](https://line-21.com/blog/2026-10-21-accessible-videos-more-options-or-more-obstacles.md) | Line 21's HLS player brings captions in many languages, optional chat and automatic delay detection, with no subscription needed for viewers to watch. | | 2026-10-14 | [Live Captioning and Profanity: Did You Miss the Punchline?](https://line-21.com/blog/2026-10-14-live-captioning-and-profanity.md) | Should live captions be allowed to swear? Why censoring profanity short-changes audiences at comedy and live shows, and what to ask your captioning provider. | | 2026-10-03 | [5 Tools for Live Captioning LinkedIn Live Streams in 2026](https://line-21.com/blog/2026-10-03-5-tools-for-live-captioning-linkedin-live.md) | Five live captioning tools for LinkedIn Live in 2026 compared on pricing and features, with guidance on which option best suits your production needs. | | 2026-09-21 | [6 Live Caption Encoding Tools for Events in 2026](https://line-21.com/blog/2026-09-21-6-live-caption-encoding-tools-for-events.md) | Compare the best live caption encoding tools for events in 2026. AV teams get pricing, CEA 608/708 support, RTMP delivery, and engine flexibility side by side. | | 2026-09-10 | [6 Livestream Captioning Tools for HLS Delivery in 2026](https://line-21.com/blog/2026-09-10-6-livestream-captioning-tools-for-hls-delivery.md) | The best livestream captioning tools for HLS delivery in 2026, with pricing, segmented caption support, latency notes and delivery options side by side. | | 2026-08-29 | [6 Live Translation Providers for Conferences in 2026](https://line-21.com/blog/2026-08-29-6-live-translation-providers-for-conferences.md) | The best live translation providers for conferences in 2026. AV teams compare simultaneous language delivery, ASR translation, human interpretation and pricing. | | 2026-08-18 | [Line 21 vs Verbit: Which Live Captioning Platform Is Right for Your AV Team?](https://line-21.com/blog/2026-08-18-line-21-vs-verbit-which-live-captioning-platform.md) | Line 21 vs Verbit for live captioning: pricing, CEA-608/708 support, RTMP delivery, human CART integration, ASR engine choice and virtual encoders compared. | | 2026-08-17 | [5 Tools for Live Captioning at Awards Ceremonies](https://line-21.com/blog/2026-08-17-5-tools-for-live-captioning-at-awards-ceremonies.md) | Planning live captions for an awards ceremony? Five tools AV teams rely on for accuracy, resilience and compliance, both in the room and on broadcast. | | 2026-08-14 | [5 Tools for Live Translation on Zoom Webinars in 2026](https://line-21.com/blog/2026-08-14-5-tools-for-live-translation-on-zoom-webinars.md) | Five live translation tools for Zoom webinars in 2026, from AI caption engines to professional human interpretation, for teams running multilingual sessions. | | 2026-08-13 | [5 Tools for Live Captioning Virtual Events in 2026](https://line-21.com/blog/2026-08-13-5-tools-for-live-captioning-virtual-events.md) | Five live captioning tools for virtual events in 2026, from compliance-focused platforms to multilingual streaming, and what AV teams should know first. | | 2026-08-12 | [5 Tools for Live Captioning at Corporate Events](https://line-21.com/blog/2026-08-12-5-tools-for-live-captioning-at-corporate-events.md) | Five live captioning tools for corporate events in 2026, from human CART accuracy to multilingual AI delivery, and what AV teams should know before booking. | | 2026-08-11 | [7 Live Captioning Tools and Platforms AV Teams Should Know in 2026](https://line-21.com/blog/2026-08-11-7-live-captioning-tools-and-platforms-av-teams-should-know.md) | Seven live captioning tools and platforms for AV teams in 2026, from human CART workflows to multi-engine ASR and broadcast delivery, and what to know first. | | 2026-08-10 | [Ai-Media vs Line 21: Live Captioning Platform Comparison](https://line-21.com/blog/2026-08-10-ai-media-vs-line-21-live-captioning-platform-comparison.md) | Ai-Media vs Line 21 for live captioning: how each handles HLS delivery, human CART support, pricing and workflow complexity, to find the right fit for you. | | 2026-08-04 | [YouTube Banned Our CTO's Account for Testing AI Dubbing. The Industry Should Take Note.](https://line-21.com/blog/2026-08-04-youtube-banned-our-ctos-account-for-testing-ai-dubbing.md) | Our CTO's YouTube account was permanently banned while testing AI speech-to-speech dubbing. A cautionary tale for AV teams working with AI audio and dubbing. | | 2026-08-03 | [7 Human CART Providers and Platforms Compared: Billing, Cancellation and Captioner Respect](https://line-21.com/blog/2026-08-03-7-human-cart-providers-and-platforms-compared.md) | Seven human CART providers and captioning platforms compared on billing transparency, cancellation terms and how they treat captioners. A guide for AV teams. | | 2026-08-03 | [Comparing Live Captioning Livestream Infrastructure Platforms: SyncWords, ENCO, Ai-Media, and Line 21](https://line-21.com/blog/2026-08-03-comparing-live-captioning-livestream-infrastructure-platforms.md) | SyncWords, ENCO, Ai-Media and Line 21 compared as live captioning infrastructure: pricing models, format support, deployment effort and human CART integration. | | 2026-08-03 | [Live Captioning for HLS Video Players: Comparing 5 Platforms for AV Teams](https://line-21.com/blog/2026-08-03-live-captioning-for-hls-video-players-comparing-5-platforms.md) | Choosing live captioning for HLS is harder than it looks. Five platforms compared on pricing, HLS architecture, human CART support and language coverage. | | 2026-07-29 | [5 Tools for Closed Captioning Across Multiple Streaming Destinations](https://line-21.com/blog/2026-07-29-5-tools-for-closed-captioning-across-multiple-streaming-destinations.md) | The best tools for closed captioning across several streaming destinations at once, compared on formats, multi-output delivery, pricing and production fit. | | 2026-07-24 | [5 Tools for Multilingual Live Caption Machine Translation at Events](https://line-21.com/blog/2026-07-24-5-tools-for-multilingual-live-caption-machine-translation-at-events.md) | Five machine translation platforms for live event captions in 2026, compared on real-time translated captions, pricing and delivery options for AV teams. | | 2026-07-24 | [How to Provide In-Room Captions at Live Events](https://line-21.com/blog/2026-07-24-how-to-provide-in-room-captions-at-live-events.md) | A practical guide to in-room captions at live events for AV teams: human captioners, ASR, audio setup, caption displays, and transparent pay-as-you-go pricing. | | 2026-07-21 | [5 Best Tools for Live Captioning on Vimeo (2026)](https://line-21.com/blog/2026-07-21-5-tools-for-live-captioning-on-vimeo-live.md) | Live captioning tools that work with Vimeo: the top five platforms for real-time captions on Vimeo Live, compared on pricing, accuracy and multilingual support. | | 2026-07-21 | [7 Best Tools for Accessible Live Event Captioning](https://line-21.com/blog/2026-07-21-7-best-tools-for-accessible-live-event-captioning.md) | Seven tools for accessible live event captioning compared: CART providers, AI captioning platforms and multilingual options for deaf and hard-of-hearing people. | | 2026-07-18 | [5 Tools for Live Captioning at Festivals and Large-Scale Live Events](https://line-21.com/blog/2026-07-18-5-tools-for-live-captioning-at-festivals.md) | Five live captioning tools for festivals and large live events: AI and human captioning, multi-stage output, pricing, and what AV teams should check first. | | 2026-07-16 | [5 Best AI Translation Tools for Corporate Town Halls](https://line-21.com/blog/2026-07-16-5-best-ai-translation-tools-for-corporate-town-halls.md) | Comparing the top AI translation tools for corporate town halls, covering multilingual captioning, enterprise integration, security compliance, and pricing. | | 2026-07-14 | [Best Tools for Live Stream Closed Captioning in 2025](https://line-21.com/blog/2026-07-14-best-tools-for-live-stream-closed-captioning.md) | The best live stream closed captioning tools for AV and production teams, compared on RTMP, HLS and CEA-608/708 support, to match your workflow and budget. | | 2026-07-10 | [5 Tools for Live Captioning on Twitch](https://line-21.com/blog/2026-07-10-5-tools-for-live-captioning-on-twitch.md) | Live captions on Twitch: five tools compared, including Line 21, AI-Media LEXI, OBS Studio, Wirecast and SyncWords, with pricing and encoder guidance. | | 2026-07-08 | [5 Tools for Real-Time Translation for Hybrid Events](https://line-21.com/blog/2026-07-08-5-tools-for-real-time-translation-for-hybrid-events.md) | Five real-time translation platforms for hybrid events compared: human interpretation, AI translation and multilingual captions for in-room and remote viewers. | | 2026-07-07 | [How to Live Caption Into Zoom via API Token Using Line 21](https://line-21.com/blog/2026-07-07-live-caption-into-zoom-via-api-token.md) | Zoom's built-in captions struggle with accents and special characters. A step-by-step guide to sending live captions into Zoom via API token with Line 21. | | 2026-07-06 | [5 Tools for Live Captioning on YouTube Live](https://line-21.com/blog/2026-07-06-5-tools-for-live-captioning-on-youtube-live.md) | Five live captioning tools for YouTube Live compared on RTMP and API delivery, CEA-608/708 encoding, multi-destination streaming, pricing and accuracy. | | 2026-07-03 | [5 Tools for Live Captioning at Keynotes and Speaker Sessions](https://line-21.com/blog/2026-07-03-5-tools-for-live-captioning-at-keynotes.md) | Five live caption tools for keynotes and speaker sessions compared, from broadcast CART to briefed human captioners and AI vocabulary training for big stages. | | 2026-06-05 | [Is Free Really Free? Answering Questions from Human Captioners](https://line-21.com/blog/2026-06-05-is-free-really-free.md) | Human captioners asked what the catch is with free Line 21 delivery. Straight answers on usage limits, data, AI training, and which services remain paid. | | 2026-05-28 | [Keeping Human Captioners in Control](https://line-21.com/blog/2026-05-28-keeping-human-captioners-in-control.md) | Line 21 is now free for human captioners delivering captions to browser, overlay, Zoom, Teams and YouTube. What's included, and why we build with captioners. | | 2026-02-11 | [Live Captioning Support: Risk vs Resilience](https://line-21.com/blog/2026-04-11-risk-vs-resilience.md) | Everyone sells live captioning support, but few define it. What real support looks like when things go wrong on show day, and the warning signs to watch for. | | 2026-01-28 | [Live Captioning: Lock-in or Loyalty?](https://line-21.com/blog/2026-03-lock-in-or-loyalty.md) | Rising captioning prices without visible value, and ecosystems that make leaving risky. How to tell if your provider earned your loyalty or just locked you in. | | 2026-01-12 | [Live Captioning Meets AI. Designing for Accuracy, Not Assumptions](https://line-21.com/blog/2026-02-live-captioning-meets-ai.md) | ASR engines and large language models do different jobs. Why mixing them into live captions risks refusals and rewrites, and how to design for accuracy instead. | | 2026-01-07 | [Captioning in 2026. A Look Back, a Look Ahead](https://line-21.com/blog/2026-01-captioning-in-2026.md) | Speech-to-speech AI and audio interpretation are on the rise, but live captioning still reaches more people. What works, and what accessibility needs in 2026. | | 2025-11-25 | [Accessible Live Captions. Talent & Technology](https://line-21.com/blog/talent-and-technology.md) | Even scripted events need a live captioner, not a technician. Why graduations, readings, panels and ceremonies rely on human skill for well-timed captions. | | 2025-11-14 | [When Captions Are Possible but Not Obvious](https://line-21.com/blog/when-captions-are-possible.md) | Captions are often available on Zoom, Teams, Webex and others, but buried in settings, permissions and plan tiers. Common restrictions and how to fix them. | | 2025-11-10 | [What We Can't Say in Live Captioning... Says the Most](https://line-21.com/blog/what-we-cant-say.md) | Why live captioning providers rarely name their clients: NDAs, confidentiality, and the paradox of building trust when you can't show off your best work. | | 2025-10-31 | [Meet the Moment: Live Captioning Display Choices](https://line-21.com/blog/meet-the-moment.md) | Browser, overlay, HLS or embedded CEA-608/708? A guide to the four main ways to display live captions, with the pros, limitations and best use for each. | | 2025-10-29 | [Live Captioning? Live Event Lessons](https://line-21.com/blog/live-event-lessons.md) | Practical lessons from captioning live events: test the real setup, secure clean audio, share context early, and plan for multiple speakers before you go live. | | 2025-10-24 | [Structuring for Scale: Live Captioning as a Foundation for Multilingual Events](https://line-21.com/blog/structuring-for-scale.md) | Live captioning as a first step toward multilingual events: how clean captions improve machine translation, support AI dubbing and help events scale up. | | 2025-10-22 | [AI and Live Captioners. A Smarter Path to Accessible Events](https://line-21.com/blog/ai-and-live-captioners.md) | AI captioning is gaining ground, but the human touch remains foundational. How live captioners can use ASR and AI to win new clients and grow their income. | --- # Resources [Documentation](https://line-21.com/docs) · [Blog](https://line-21.com/blog) ## For AI agents This website is written to be read by AI assistants and agents as well as people. Point your tools at these; they are generated from the same content as the pages, so they stay in sync. - [/llms.txt](https://line-21.com/llms.txt): An index of the whole site for language models: what Line 21 is, with links to every page in Markdown. - [/llms-full.txt](https://line-21.com/llms-full.txt): The main pages and all the documentation in a single Markdown file. - Any page + .md (e.g. [/pricing.md](https://line-21.com/pricing.md)): Every page has a Markdown version: add .md to its address, or request it with an Accept: text/markdown header. - [/languages.json](https://line-21.com/languages.json): Every supported language with its human captioning, speech recognition, translation and AI dubbing coverage. - [/pricing.json](https://line-21.com/pricing.json): Plans, support levels, pay-as-you-go rates, and the plan builder's pricing model. ## Frequently asked questions ### Writer Functions **How do I connect to Line21?** Three methods: use our connector app (download from above), use macro to write directly into a text box (similar to how you would write in a Google Doc), or you can opt for integrating with our API. **How do I start and stop my session?** You don't - Line21 will track your active captioning minutes for you and you will be charged only for minutes used. [See more about billing here](https://line-21.com/docs/getting-started/organisations#billing) . **What if I have a long session and need a co-worker?** Co-working is simple - invite your other writers to your link, and work in tandem. **Can I download transcripts at the end of the session?** Absolutely - as .TXT, .SRT, and .VTT. You can download your translated languages as well. Yes, we support [AI Translations](https://line-21.com/docs/project-management/assets) as well. ### Viewer Options **What about encoding?** You can video encode too and simultcast to several RTMP endpoints! Under [destinations](https://line-21.com/docs/destinations/rtmp-streaming) check for 607/708 encoding. **What about integrations with video-conferencing platforms?** We are able to integrate via API with all the standard platforms: Zoom, YouTube, MS Teams. If we don't have it listed, [let us know](https://line-21.com/contact) and we'll have it in record time! **What options do I have for captions?** For basic virtual functions: a browser link, a subtitle overlay, and an iframe. We consider these “destinations” and you can enable other destination such as video-conferencing platforms, live events platforms and even video encoding. Check [documentation on destinations](https://line-21.com/docs/destinations) . **Will viewers be able to customise their captions?** Viewers can either choose one of the presets we offer, opt for the style you can customise for the entire event, or they can simply customise a style that they prefer: fonts, sizes, colours, weights, etc. [See caption style customisation page.](http://localhost:3000/docs/destinations/browser#custom-style **Is Line21 screen reader friendly?** It is! Line 21 is entirely web-based and our components are WCAG compliant. We are using Mantine UI which is one of the most reader-friendly UI libraries. ### Translations **Can I offer Translations?** We provide translations in 111 languages. We use the best engines in the market, to ensure accuracy and stay ahead of the curve. [See all about AI Translations](https://line-21.com/docs/project-management/assets) . **How do other language writers fit in?** We treat it the same as normal co-working. Invite the writers to the same link. Each writer will be able to choose their language, and contribute captions in that language. Line21 will detect the language input and allocate to the correct output for the viewer. ### AI Proofreader **How does the AI Proofreader work?** Once you enable the AI Proofreader you can provide context information to prepare it. You have a choice of three trust modes and two speed settings. [Read all about it here](https://line-21.com/docs/ai-services/ai-proofreader) . **How do I prepare the AI Proofreader?** You feed context glossary and terms. We encourage you to [talk to us](https://line-21.com/contact) before first use to ensure you're maximizing its potential. ### Billing **How does billing work?** There's a basic pay-as-you-go plan. You pay for minutes used at month end on a fixed rate. You can also buy bundles of minutes upfront to optimise cost (up to 50% off standard price). [Read about minutes and billing](https://line-21.com/docs/getting-started/organisations#billing) . **How does billing work for premium features - AI Translation & Proofreader?** You will be invoiced separately at the end of each month, for all active minutes of either service used. [See billing for current rates](https://line-21.com/docs/getting-started/organisations#billing) . [Contact us for special use cases](https://line-21.com/contact) . **Do you offer different sized bundles?** Absolutely - and talk to us if you want a different package. [See packages here](https://line-21.com/docs/getting-started/organisations#billing) . ### Technical **Is Line21 compatible with all browsers?** Yes, we support all modern browsers. Best experience on Google Chrome, Microsoft Edge, Mozilla Firefox and other browsers based on these. Including Safari. [If you have a special use case, contact us](https://line-21.com/contact) . **How secure is Line21?** We employ industry-standard security procedures and use secure cloud providers such as Vercel and Microsoft Azure. For data retention, AI training, third-party processors, and client-configurable options (EU residency, ZDR, vendor exclusion), see our [data handling & retention overview](https://line-21.com/docs/security/data-handling) . **How stable / scalable is Line21?** We've built Line 21 on an architecture that allows for a virtually endless scalability. If you have a particularly large event and you want to discuss further, [contact us](https://line-21.com/contact) . **Do I need the internet?** Yes, Line21 is an internet based software for virtual and hybrid audiences everywhere. You can use Line 21 for in-person events, but you still need internet to make it work. We distribute, translate and proofread captions automatically using a network of services that require internet access. ### Additional Questions **What is the difference between closed and open captions?** Closed captions can be turned on and off in a video player. Open captions cannot be turned off. They are usually burned into the video itself. **How can I upload my transcript to another platform later?** You can download your transcript as an .SRT or .VTT file and upload it to your chosen platform. **Do you support onsite, offline captioning?** At this time, we are not providing an offline platform. Line 21 requires internet to operate, even when used for an in-person event. **Do you support subtitle editing?** We expect to add this to our future features in 2024. **Do you support respeaking?** At this time, no - though we are actively working on a solution that will allow respeakers to work directly in our platform without other software solutions required. Stay tuned - subscribe to our newsletter here. Or [contact us if you want to know more](https://line-21.com/contact) . **Do you support sign language models?** At this time, no - though this will also change in the future. [Contact us if you want to chat about this](https://line-21.com/contact) . --- # About us Line 21 is AI infrastructure for live media language services: captions, translation and AI dubbing in 100+ languages, for any live event. - **Founded**: 2023 - **Team**: 3 - **Languages**: 100+ - **Backed by**: SFC Capital ## What we do AI made multilingual live events affordable, and now organisers want them in ten languages at once, with captions and dubbing, on every platform. Generic AI stumbles on names and jargon, and nothing joins the pieces together. Line 21 orchestrates speech recognition, translation and voice models alongside human captioners, and delivers the result to venue screens, streams and meeting platforms in real time. Two technologies sit at its core: an AI Proofreader that checks every caption line against a glossary built for each event, and a cloud video encoder that puts captions into live streams on any platform. ## Team A deliberately lean team, led by founders who have been building products for more than eight years. - **Paul Balogh**, CEO, Co-founder: Tech entrepreneur and frontend developer, founder of Zeta Books, Learn Forward and Hypersay. https://www.linkedin.com/in/paulbalogh/ - **Fabrizio Ruggeri**, CTO, Co-founder: Senior full-stack developer and solutions architect. Worked for Hypersay and PrestaShop. https://www.linkedin.com/in/fabrizioruggeri/ - **Victoria Hart**, Head of Growth: Extensive industry knowledge from working for a major UK-based captioning agency. https://www.linkedin.com/in/thatvictoria/ ## Backed by SFC Capital Line 21 is venture-backed by SFC Capital, the UK's most active seed investor. https://sfccapital.com/ ## Investor enquiries Interested in Line 21's growth? Write to Paul Balogh, our CEO and co-founder, directly. paul@line-21.com --- # Contact us Questions about captions, plans for an event, or help on the day: talk directly to the team that builds and runs Line 21. - **Email**: hello@line-21.com - **Live chat**: on any page of [line-21.com](https://line-21.com/) - **Book a demo**: https://cal.com/victoria-line21/lets-talk - **Get a quote**: the form on [the contact page](https://line-21.com/contact) (name, company, email, services, event date, languages, message) - **Discord**: https://discord.gg/ZgCYyXywCE --- # Zoom integration ## Why Line 21 Live Captions? Our simple to use app delivers a higher accuracy automated captioning service for attendees in Zoom meetings and webinars, either in the original language, or supporting machine translation. ## Registering for a Line 21 Account [https://scribehow.com/embed/Registering_your_Line_21_account__-BLuI3t4QO2CCi0T90_-lw](https://scribehow.com/embed/Registering_your_Line_21_account__-BLuI3t4QO2CCi0T90_-lw) Once your account and card is registered, you can follow our step by step guide for using the app below. --- ## **Using the Line 21 Live Captions App** A step-by-step guide for users to add Line 21 Live Captions to their Zoom Meeting, made with Scribe. [https://scribehow.com/embed/Create_Live_Captions_for_Zoom_Meeting__lCT_13sJQY2QT60hXQp67Q](https://scribehow.com/embed/Create_Live_Captions_for_Zoom_Meeting__lCT_13sJQY2QT60hXQp67Q) --- ### **Removing the App** To remove our Line 21 Live Captions app from your Zoom account, you can take the following steps: 1. Login to your Zoom Account and navigate to the [Zoom App Marketplace](https://marketplace.zoom.us/). 2. Click **Manage** >> **Added Apps** or search for the "Line 21 Live Captions" app. 3. Click the "Line 21 Live Captions" app. 4. Click **Remove**. We’re sorry to see you go, and hope you’d like to share your reasons with us to improve our service - please email [hello@line-21.com](mailto:hello@line-21.com) with any feedback you may have. --- ## Troubleshooting **My Zoom bot won’t join the call?** Check your Zoom invitation is correct. **My captions aren’t appearing on the screen?** Check you have created a Zoom destination in the Line 21 dashboard; if missing, please create one. Next, please check you have provided the correct API token in Line 21 ‘Destination’. If the option for providing an API token is not appearing, please check your Zoom meetings and enable manual captions as per [Zoom’s guidance](https://support.zoom.com/hc/en/article?id=zm_kb&sysparm_article=KB0060368). --- ### **Contact Support** All support requests will be responded to within 24 hours - we ask our users to email [hello@line-21.com](mailto:hello@line-21.com). --- # Terms of service Questions: legal@line-21.com ## Terms of Service **Line21** is a service that includes Line21 services provided to you and all persons or entities using your Account, subject to the terms of this Agreement. Users who violate the terms of this Agreement will be immediately terminated from access to the Website and will have their account canceled. Scheduling an event or uploading a document or saving a document within Line21 constitutes your complete acceptance of the terms and conditions of this agreement. If you do not agree to abide by these terms, you will not be permitted to use the Line21 service. ## Intellectual Property You agree that you will not transmit or upload any Content to the Website that infringes any patent, trademark, trade secret, copyright or other proprietary rights ("Rights") of any party. You may not use any trademark or service mark appearing on the Website without the prior written consent of the owner of the mark. You agree that you will not re-engineer or copy any of the functions or features of the Line21 platform. ## Content You and any user of your Account must evaluate, and bear the risk associated with, the accuracy, completeness or usefulness of any Content. Line21 shall not be responsible for any Content broadcast through the Line21 platform to any and all RTMP destinations. Line21 does not pre-screen or monitor all Content as a matter of policy, but Line21 shall have the right, but not the responsibility, to remove Content which is deemed in the sole discretion of Line21 harmful, offensive, or otherwise in violation of this Agreement or any rules that Line21 has in place at the time. Line21 may elect at its sole discretion to monitor some, all, or none of the events for adherence to this Agreement or the rules set out by Line21. ## Disclaimer of Warranties You expressly agree that use of Line21 and the Website is at your own risk. Line21 and all affiliates shall not be responsible for any content found on the Website or broadcast using Line21 services. Line21 and all affiliates expressly disclaim all warranties of any kind, whether express or implied. Additionally, Line21 and all affiliates make no warranty that: 1. The Line21 streaming services or website will meet your requirements; 2. The Line21 streaming services or website will be uninterrupted, timely, secure, or error free; 3. Any results may be obtained from the use of the Line21 streaming services or website; and/or 4. Any information obtained through the Line21 services or website will be accurate or reliable. 5. Any results may be obtained from the use of theLine21 services or website; and/or 6. Any information obtained through the Line21 website will be accurate or reliable. No advice or information, whether oral or written, obtained by you from the Line21 services or website shall create any warranty not expressly made herein. You understand and agree that any content downloaded or otherwise obtained through the use of the Website is done at your own discretion and risk and that you will be solely responsible for any damages to your computer system or loss of data that results in the download of such content. You understand and agree that copyright laws protect any copyrighted material that you download or otherwise obtain through the use of the Website and regulations and that downloading and/or using copyrighted material in an illegal manner could result in lawsuits, damages, or other sanctions against you. ## Limitation of Liability To the fullest extent permissible under the applicable law, Line21 and all affiliates shall not be liable for any damages arising out of the use of the Website service, including any damages from viruses alleged to have been obtained from the Website service and including any direct, indirect, incidental, special or consequential damages resulting from the use or inability to use the Website, or for the cost of procurement of substitute goods and services, or resulting from any goods or services purchased or obtained or messages received or transactions entered into through the Website, or resulting from unauthorized access to or alteration of profits, use, data, or other intangible properties, even if Line21 and all affiliates have been advised of the possibility of such damages. ## Termination of Account You or Line21 may terminate your Account at any time for any reason. Once your Account is terminated, you cannot access the Website in any manner or for any reason, and you cannot assist in enabling a terminated user access to the Website using or via your Account. ## Other account users In cases where you have authorized or registered another individual, to use your Account, you recognize and agree that you are fully responsible for: 1. The online conduct of such user at the Website; 2. Controlling the user's access to and use of the Website; and 3. The consequences of any misuse. You further recognize and agree that the holder of the Account is entirely responsible for all activities conducted through such Account. ## Miscellaneous Line21 reserves the right to revise this Agreement at any time, and you should periodically review this Agreement to determine the conditions to which you are bound. The opinions expressed by users in any of the events are not necessarily those held by Line21. --- # Privacy policy At Line21, safeguarding the privacy and data security of our users is a top priority. This Privacy Policy outlines our commitment to transparently handle user information while ensuring a personalized and secure experience. Please read this policy carefully to understand how we collect, use, and protect your data. ## Information We Collect We collect the following minimal amount of user information necessary for enhancing your browsing experience and facilitating event interactions: 1. **User's Name**: We use your name to personalize your interactions with Line21 services. 2. **User's Email Address**: Your email address enables us to communicate important updates and allows you to invite others to Line21 events. 3. **Transcript of Line21 Event**: By default, Line21 does not retain transcript data. However, for users who explicitly choose this option when scheduling an event, transcript data will be securely stored. ## Data Usage and Monitoring Discussions and activities conducted through Line21 events may be monitored as deemed necessary by Line21. This helps us maintain the quality and security of our services. ## Data Disclosure Line21 is committed to not disclosing user communication contents unless required by law or under circumstances where such disclosure is reasonably necessary to: 1. **Comply with Legal Process**: We may release information if required to abide by legal procedures or comply with law enforcement requests served on Line21. 2. **Protect Our Rights**: We may disclose information to protect and defend the rights and property of Line21 or to uphold the terms of a subscriber's agreement. 3. **Ensure Safety**: We may release information to ensure the personal safety of Line21 participants. ## Data Analysis and Retention Line21 does not engage in data analysis for third-party sale purposes. We also do not perform any form of data analysis on retained data. Data retained is solely for the benefit of Line21 account holders and their event-related needs. ## Your Choices You have the option to control your data by selecting whether transcript data is retained for your events. This choice underscores our commitment to providing you with tailored control over your data. ## Contact Us If you have any questions or concerns regarding our Privacy Policy, please reach out to our dedicated privacy team at [hello@line-21.com](mailto:hello@line-21.com). ## Effective Date This Privacy Policy is effective as of 14th August 2023 and supersedes any previous versions. We reserve the right to amend this policy and will notify users of any substantial changes. Your trust is valuable to us, and we are committed to ensuring your privacy while delivering an exceptional user experience at Line21. ### Right to Access, Change, or Delete Personal Data 1. Right to Access. Individual Customers have the right to know what Personal Data about them is included in the databases and to ensure that such Personal Data is accurate and relevant for the purposes for which Line21 collected it. Individual Customers may review their own Personal Data stored in the databases and correct, erase, or block any data that is incorrect, as permitted by applicable law and Line21 policies. Upon reasonable request , Line21 allows Individual Customers access to their Personal Data, in order to correct or amend such data where inaccurate. Individual Customers may edit their Personal Data by logging into their account profile or by contacting Line21 by phone or email. In making modifications to their Personal Data, Data Subjects must provide only truthful, complete, and accurate information. To request erasure of Personal Data, Individual Customers should submit a written request to the Line21 office. 2. Requests for Personal Data. Line21 will track each of the following and will provide notice to the appropriate parties under law and contract when either of the following circumstances arise: (a) legally binding request for disclosure of the Personal Data by a law enforcement authority unless prohibited by law or regulation; or (b) requests received from the Data Subject. If Line21 receives a request for access to his/her Personal Data from an Individual Customer, then, unless otherwise required under law or by contract with such Individual Customer, Line21 will refer such Data Subject to the Individual Customer. 3. Satisfying Requests for Access, Modifications, and Corrections. Line21 will endeavor to respond in a timely manner to all reasonable written requests to view, modify, or inactivate Personal Data following an official request to do so. ### Cookie Policy Line21 does not use cookies at this time. ### Your Data Requests and Rights We are committed to addressing requests from individuals seeking to exercise their data protection rights in accordance with relevant data protection laws. In order to safeguard your privacy and security, we may need to verify your identity through reasonable measures before responding to your request. To initiate any of these rights, feel free to reach out to us at [hello@line-21.com](mailto:hello@line-21.com). If you find that your concern remains unresolved, you have the option to contact the data protection authority of your country. You might also possess the right to file a complaint with the appropriate Supervisory Authority under GDPR regulations. A comprehensive list of Supervisory Authorities can be accessed at: [http://ec.europa.eu/justice/data-protection/bodies/authorities/index_en.htm](http://ec.europa.eu/justice/data-protection/bodies/authorities/index_en.htm). Should you require further assistance regarding your rights, kindly get in touch with us at [hello@line-21.com](mailto:hello@line-21.com). We will diligently consider your request in alignment with applicable laws. It's important to note that in certain instances, our capacity to honor these rights may be contingent upon our obligations to process personal data for purposes such as security, safety, fraud prevention, regulatory compliance, legal obligations, or the provision of requested services. In such cases, we will furnish you with pertinent details relevant to your request. ### Data Transfers to the United States The Line21 platform may process and store information in the U.S. Therefore, we and our service providers may store and access your personal information in the U.S. The U.S. may not provide equivalent levels of data protection as enjoyed in your home jurisdiction. It may occur that a third-party based in the EU, such as your employer, may need to transfer personal information to us for the purpose of verifying your participation in a class. In such instances, whenever your personal data is transferred to countries outside of the EAA, UK or Switzerland, we will ensure that at least one of the following safeguards is in place: - The country is one that the European Commission have approved as providing an adequate level of protection for personal data; - The transfer is subject to a specific derogation in the GDPR or national laws; - Through the use the standard contractual clauses as the transfer mechanism when a case-by-case analysis has been performed; or - Where we use certain service providers, we may use specific contracts or codes of conduct or certification mechanisms approved by the European Commission which give personal data substantially similar protection as in the UE, EEA or UK. ### How you can access, update or delete your data. Ensuring the accuracy and currency of the information we have about you is crucial. If your personal information undergoes any changes or if you wish to opt-out of email communications, please don't hesitate to reach out to us. You can drop us an email at: [hello@line-21.com](mailto:hello@line-21.com). We're here to assist you at any time. --- # API (/docs/api) ## Overview [#overview] Jump to the [full API documentation](https://caption.line-21.com/api/v1) Line 21 API enables developers to interact with the system programmatically. The platform provides a REST API which can be interacted with through HTTP and from any programming language with all requests and responses encoded in JSON format. The API currently supports two actions with other expanded functionality planned for [future releases](/docs/api#future-plans): * **Caption production**: add external captions to a project * **Caption retrieval**: download recorded captions for a project in various formats ## Authentication [#authentication] API Key Screenshot Authentication is based on the usage of a `Bearer Token` linked to a project.\ In your project dashboard navigate to `API` in the left sidebar and copy the `API Key`. Each request must contain an `Authentication` header with `Bearer ` as content. Note: the API key is unique to the project not to the organisation. This is useful especially if you work with different partners or clients for different projects. ## Endpoints [#endpoints] Send your requests to `https://caption.line-21.com/api/v1` with JSON-encoded payloads.\ For comprehensive endpoint documentation, request/response schemas, and interactive examples, visit the complete [API documentation](https://caption.line-21.com/api/v1). ### Example request [#example-request] You can see an example using cURL from command line ```bash curl -X POST "https://caption.line-21.com/api/v1/captions" \ -H 'accept: application/json'\ -H 'authorization: Bearer '\ -H 'content-type: application/json' \ -d '{"projectId":"prj-id-123","language":"en","text":"The lazy fox...","timestampStart":1687344200000,"timestampEnd":1687344250000,"isInterim":false}' ``` ## Future plans [#future-plans] We're actively expanding the API to provide complete programmatic control over the Line 21 platform. Upcoming features include: * Organization and project management endpoints * Organization-level API keys with granular permissions * Multi-key support for enhanced security and access control * Full coverage of all platform capabilities [Contact us](https://line-21.com/contact) if you have any questions or feedback about the API, including requests for new endpoints or features. --- # FAQs (/docs/faqs) Welcome to our FAQs — your go-to resource for quick answers to common questions about our services, team, and workflows. This is a living document that is updated frequently - if you have a question and can't see it here, please email [hello@line-21.com](mailto:hello@line-21.com) and we'll add it to our knowledge base! ## General FAQs [#general-faqs] **What options do I have for captions display?** * Traditional browser * Captions Embed * Subtitle overlay * Text destinations through APIs (e.g. YouTube, Zoom, etc.) * RTMP / CEA-608/708 Closed Captions * HLS Video Closed Captions **Can viewers customise their captions?** Viewers can either choose one of the presets we offer, opt for the style you can customise for the entire event, or they can simply customise a style that they prefer: fonts, sizes, colours, weights, etc. [See the caption style settings.](/docs/destinations/browser#settings) **Is Line21 screen reader friendly?** It is! Line 21 is entirely web-based and our components are WCAG compliant. We are using Mantine UI which is one of the most reader-friendly UI libraries. **What is the difference between closed and open captions?** It's a bit counterintuitive. Closed captions are the ones that can be turned on and off in a video player (look for the CC button). Open captions cannot be turned off. They are usually burned into the video itself. **Do you support onsite, offline captioning?** We can provide onsite support for live events, but would require a hardwired internet connection for our innovative software to support our human captioners. At this time, we are not providing an exclusively offline platform. Line 21 requires internet to operate, even when used for an in-person event. **Do you provide professional captioners?** Yes, all of our human captioners are accredited and registered with the relevant organisation in their country of origin. We work with stenographers and voice writers, as methods will vary depending on language. ## CART Captioning FAQs [#cart-captioning-faqs] **How do I connect to Line 21?** The main connection method is via writer link; via keyboard macro or Stenokeys to output text on mouse cursor. If you are an **Eclipse** user, you can download our connector from your Line 21 dashboard. If you are a **CaseCAT** user, we invite you to open a ticket with Stenograph to support our progress in integrating with them. **How do I start and stop my session?** You don't - Line21 will track your active captioning minutes for you and you will be charged only for minutes used. **How can I work with a co-worker?** Co-working is simple - invite your other writers to your link, and work in tandem. The system will take text from either writer. **Is my writer page what the viewer sees?** No, viewers have a different output link that resembles traditional browser output. This is available in your 'destinations' section in the dashboard. **How do I send a new line of text?** In the Project dashboard, either yourself or the project creator can change the settings of line output to automatically send on: * New Line / Enter * Character length * Punctuation If you prefer to keep control in your hands, then you will need to use a keyboard macro to push captions on 'Enter' with a tool like Stenokeys. **Does Line 21 support respeaking?** Absolutely! We offer Dragon users a specific respeaker input page, fine tuned to counter typical respeaking issues such as 'dragging'. ## Security & data [#security--data] **Is client content used to train third-party AI models?** No. Client content processed through the configured Line 21 pipeline is not used to train third-party AI models. See the full [Data handling & retention](/docs/security/data-handling) article for vendor-by-vendor commitments. **Can specific vendors be excluded from our pipeline?** Yes. Vendor selection can be constrained per client engagement where policy requires it. **Is EU data residency available?** Yes, depending on the selected speech-recognition and translation providers. EU endpoints are available for selected ASR and translation providers, including Deepgram EU and regionalised translation endpoints. **Is Zero Data Retention (ZDR) available?** Yes. ZDR agreements are available for OpenAI and Anthropic engagements that require them. **Are HLS recordings retained?** By default, HLS recordings are retained for 30 days on Line 21 infrastructure. This can be shortened or configured per client engagement. See [HLS streaming](/docs/destinations/hls-streaming) for setup details. --- # Overview (/docs) ## Essentials [#essentials] The core concepts you’ll use in almost every setup: how captions are created, how languages work, and how you deliver them to viewers. - [Getting started](/docs/getting-started): Create an account, set up your organisation, and learn the basics. - [Captioning methods](/docs/methods): Human, ASR, and playback — when to use each and what to expect. - [Languages](/docs/languages): How language channels work (and how to add translations). - [Destinations](/docs/destinations): Where captions go: browser, overlay, streaming, and text destinations. *** ## Advanced [#advanced] Deep dives for production workflows, integrations, and teams that want more control. - [AV feeds](/docs/feeds): Understand audio/video inputs and how they affect caption quality. - [AI services](/docs/ai-services): Add translations, proofreading, and other assistive capabilities. - [Project management](/docs/project-management): Monitor live events, coordinate with the team, and review assets after. - [API](/docs/api): Programmatic integrations for caption production and retrieval. - [Line 21 app](/docs/line-21-app): Operational how‑tos for running a live event in the web app. - [Security & data](/docs/security/data-handling): Data retention, AI training, third-party processors, and client options. - [FAQs](/docs/faqs): Quick answers to the questions we hear most often. *** For any questions, feedback, or support, [please contact us here](https://line-21.com/contact). --- # AI Call Agent (/docs/ai-services/ai-call-agent) ## When to use AI call agent [#when-to-use-ai-call-agent] Line 21 is able to transcribe your meetings through an AI call agent (sometimes referred to as a **Bot**). The Bot can automatically join your calls and transcribe the audio using the powerful setup provided by [Line 21 ASR](/docs/methods/asr). ## Call agent vs Meeting project type [#call-agent-vs-meeting-project-type] Line 21 has two features that are useful for transcribing meetings: | Feature | Description | When to use | | --------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------- | | [Meeting project type](/docs/project-types/meeting) | A project type that is designed for meetings. Participants join a special page with ASR enabled in their language. The captions are cross-translated to all other languages. | Multilingual conference calls with minimum of two source languages where strict diarisation is important. | | AI Call Agent | Transcribe entire meeting by having an ASR bot joining the call. Translate the transcript in real-time to any languag. | Single or multi-language conference calls with limited diarisation. | ## Apps supported [#apps-supported] | Platform | Status | | --------------- | --------------------- | | Zoom | 👍 Supported | | Microsoft Teams | 🛠️ Under development | | Google Meet | 🛠️ Under development | ## Zoom [#zoom] ### Prerequisites [#prerequisites] * A Zoom account and sufficient permissions to add bots to your meetings. * A Line 21 standard project with at least one primary lanaguage enabled in [ASR mode](/docs/methods/asr). ### Set up your Zoom account [#set-up-your-zoom-account] Setting up a Bot for Zoom is easy. First, add our integration to your account through the\ [Zoom Marketplace](https://marketplace.zoom.us/apps/_wzvvDgGQUe2OPkMa1vtDQ). ### Set up your Line 21 bot [#set-up-your-line-21-bot] In your Line 21 project, in the **AI Call Agent** section, click on the plus icon to add a new bot. Add bot Fill in the following information: 1. **Bot Name** — The name shown to participants in the call. 2. **Platform** — Only Zoom is available at the moment. 3. **Join Meeting URL** — Your meeting URL. If the meeting was scheduled in advance, use the same link you shared with participants. Zoom Webinar links are also supported. 4. **Primary Language** — The language spoken in the meeting, as previously configured in the project. After saving, you will see a new AI Call Agent entry in your project. ### Run a Zoom Bot [#run-a-zoom-bot] Once your Zoom Bot is ready, you can ask it to join your call by clicking the **Join call** button. Add bot When clicked, the Bot will attempt to join the call. The host may need to admit the Bot depending on your Zoom settings. As soon as the Bot joins, it will request permission to record the call. This is required for the Bot to access audio. Line 21 does **not** store any audio from your meetings — it is only used live to perform automatic speech recognition. Once the host grants recording permission, the transcription will start. ### Show captions on Zoom [#show-captions-on-zoom] If you want the transcription to appear directly inside the Zoom call, you can configure a\ [Zoom API destination](/docs/destinations/text-destinations#zoom-api) (supports only one language). Alternatively you can share a [browser link](/docs/destinations/browser) to the participants (supports all configured languages & translations). --- # AI Context (/docs/ai-services/ai-context) AI services, much like humans, work best when they have a clear understanding of the context of the event. For humans, we are talking about `preparation materials`. For AI Services (like ASR, or the AI Proofreader), it is called `AI Context`. The ultimate goal is to increase overall live captioning accuracy. ## What does it look like? [#what-does-it-look-like] Context can be anything from: * Scripts * Slide decks * Run of show / agendas * Names; whether speakers, place, brand, or element * Stats and figures; any mission critical information that must be spot-on * Homophones; is it Sara or Sarah? Sysco or Cisco? * Accents or dialects in place Contexts can be as short or long as you want. Even a short context simply indicating the broad domain is proved to be effective. Examples: The event is about molecular biology. The event is about molecular biology, specifically the study of DNA sequencing and its applications in medicine. The conference title is "Double Helix and Beyond". The event is hosted by the company "Guillermo DNA Sequencing Inc." and the keynote speakers are Dr. Cisco Frank and Dr. Sara Enstein. The event is about molecular biology, specifically the study of DNA sequencing and its applications in medicine. The conference title is "Double Helix and Beyond". The event is hosted by the company "Guillermo DNA Sequencing Inc." and the keynote speakers are Dr. Cisco Frank and Dr. Sara Enstein. The abstract: "Recent advances in high-throughput DNA sequencing have reshaped our capacity to read, interpret, and algorithmically recompose biological information at scales previously relegated to speculative fiction. Leveraging long-read nanopore platforms and error-corrected synthetic assembly, researchers are now able to reconstruct entire developmental gene circuits with a temporal precision approaching single-cell lineage resolution. Here we review emerging methodologies that fuse sequencing-derived regulatory maps with ex vivo biofabrication workflows—an interface sometimes referred to, half-jokingly and half-nervously, as “neo-Frankenstein” biotechnology." ## How is it used? [#how-is-it-used] | Service | Description | Updates | | -------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------- | | [AI Proofreader](/docs/ai-services/ai-proofreader) | Makes maximal use of the global context provided. It will use the `Description`, `Names`, and `Specialised Language` to proofread the text in real-time. | Continuously while the event is running | | [ASR context](/docs/methods/asr) | ASR engines use a derived context in the form of a `dictionary`. Some models allow for `sound like` functionality, which is a way to tell the ASR model how to expect a certain word to be pronounced, and how to spell it correctly. | On starting the ASR engine | ## Managing Context [#managing-context] Line 21 offers a chat-based tool for setting up and maintaining your global context called the AI Context Builder. ### Creating Context [#creating-context] AI Context Builder By interacting with the AI Builder, you can: * Set up a global context for your event by simply chatting with the AI Context Builder and telling it about your event: * You can paste large amounts of structured or unstructured text (copy / paste a document or a website) * You can add urls for the chat to parse * You can upload documents to be parsed * You can modify the global context at any time The AI Context Builder will parse the information you provide, extract relevant information following 3 main categories: | Type | Description | | ---------------------- | -------------------------------------------------------------------------------------------------------------------------------- | | `Description` | a general description of the event (can be as short or long as you want). | | `Names` | a list of proper nouns that should be recognized with high accuracy (speakers, authors, companies, product names, places, etc.). | | `Specialised Language` | any highly specific language, jargon, or terminology that should be recognized. | You can always switch to `Manual` mode, which is a human-curated context by turning off the `Auto-Update` toggle. Use this when you want to fine-tune the generated context or the AI Builder is not working as expected. For some events, context is easier to review by downloading it - this can be done by using our download context feature. This is also a helpful tool for our human captioners, saving them time parsing through event information. ### Updating the Context [#updating-the-context] While you can update the global context at any time, you should be aware of how this affects the derived contexts for the ASR and the Proofreader. * The **AI Proofreader context** is updated automatically when the global context is updated. It will produce effects from the very next proofread caption without the need for any further action. * The **ASR Context** will be updated automatically, *but it will not take effect until the next ASR engine restart*. ASR Contexts are being activated once at the beginning of an ASR session. Resetting the ASR connection will activate the new context. ### Clear Context [#clear-context] Choose `Clear Context` to start fresh - hold the red button on the right, and current context will be deleted. --- # AI Dubbing (/docs/ai-services/ai-dubbing) AI Dubbing turns the speaker's voice into spoken audio in another language, live. Your audience can listen in their own language instead of reading captions. AI Dubbing is in **beta** and is switched on per organisation. [Contact us](/contact) to enable it. In the dashboard it appears as **Interpretation**, under AI Agents in your project settings. ## How it works [#how-it-works] AI Dubbing uses a **speech-to-speech** model: the original audio goes in, translated speech comes out, with no captions in between. We currently use one model: **Google Gemini Live Translate**. It works on a continuous stream, translating while the speaker is talking rather than waiting for them to finish a sentence. * **Language changes**: the model follows a live stream that switches language, without restarting. * **Voice matching**: the dubbed voice follows the original speaker, including their gender. * **In sync**: every dubbed track stays aligned with the original video. ## Supported languages [#supported-languages] AI Dubbing is available in 75 languages, listed below and on the [Languages](/languages) page. The source must be a single-language input: inputs that combine several languages can't be dubbed yet. | Code | Language | | --- | --- | | AF | Afrikaans | | SQ | Albanian | | AM | Amharic | | AR | Arabic | | HY | Armenian | | AZ | Azerbaijani | | EU | Basque | | BE | Belarusian | | BN | Bengali | | BG | Bulgarian | | CA | Catalan | | ZH | Chinese (Simplified) | | HR | Croatian | | CS | Czech | | DA | Danish | | NL | Dutch | | EN | English | | ET | Estonian | | FI | Finnish | | FR | French | | GL | Galician | | KA | Georgian | | DE | German | | EL | Greek | | GU | Gujarati | | HA | Hausa | | HE | Hebrew | | HI | Hindi | | HU | Hungarian | | IS | Icelandic | | ID | Indonesian | | IT | Italian | | JA | Japanese | | JV | Javanese | | KN | Kannada | | KK | Kazakh | | KM | Khmer | | RW | Kinyarwanda | | KO | Korean | | LO | Lao | | LV | Latvian | | LT | Lithuanian | | MK | Macedonian | | MS | Malay | | ML | Malayalam | | MR | Marathi | | MN | Mongolian | | MY | Myanmar (Burmese) | | NE | Nepali | | NO | Norwegian | | FA | Persian | | PL | Polish | | PT | Portuguese | | PA | Punjabi | | RO | Romanian | | RU | Russian | | SR | Serbian | | SD | Sindhi | | SI | Sinhala (Sinhalese) | | SK | Slovak | | SL | Slovenian | | ES | Spanish | | SU | Sundanese | | SW | Swahili | | SV | Swedish | | TL | Tagalog (Filipino) | | TA | Tamil | | TE | Telugu | | TH | Thai | | TR | Turkish | | UK | Ukrainian | | UR | Urdu | | UZ | Uzbek | | VI | Vietnamese | | ZU | Zulu | ## Adding AI Dubbing [#adding-ai-dubbing] 1. Open your project's **Settings** and find **Interpretation** under **AI Agents**. 2. Click **+** to add an interpretation. 3. Choose the **Model** (Gemini Live Translate). 4. Choose the **Input source**: the language input whose speech you want to dub. 5. Choose the language to dub into, then click **Add Interpretation**. The interpretation shows as **Idle** until audio is flowing, and you can add one for each language your audience needs. ## Where the dubbed audio goes [#where-the-dubbed-audio-goes] | Destination | How the dubbing is delivered | | --------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------- | | [HLS streaming](/docs/destinations/hls-streaming) | Each dubbed language is published as an alternative audio track that viewers pick in the player. | | [RTMP streaming](/docs/destinations/rtmp-streaming) | The original audio is replaced with one dubbed track. Turn off **Use original audio** and pick the **Audio track**. | | SRT | Dubbed audio from existing interpretations can be muxed into the SRT output. | | Captions page | Coming soon: listen to the dubbed audio on the same page as the captions. | ## Things to know [#things-to-know] The model's voice replication is good but not perfect. As Google notes, a voice can occasionally shift after a long pause, start with the wrong gender, or settle on one voice when several people talk in quick succession. Language detection can also struggle with heavy accents or very similar languages, although the translation itself should still be accurate. For a high-stakes session, test with a recording of your speakers first, and keep captions available alongside the dubbed audio. Dubbed audio is processed by the same pipeline as the rest of your content: see [Data handling & retention](/docs/security/data-handling). --- # AI Proofreader (/docs/ai-services/ai-proofreader) ## What's the AI Proofreader? [#whats-the-ai-proofreader] The AI Proofreader is an LLM-powered AI safeguard layer that, based on the [generic AI context](/docs/ai-services/ai-context), will proofread the captions lines in real-time. It works with both methods of producing captions ([human](/docs/methods/human) and [ASR](/docs/methods/asr)). It works best with events where the language used is highly specific and technical or where correct spelling of brands and names is critical. ### What it does [#what-it-does] The AI Proofreader will dynamically supervise any captions lines and apply the following checks: * Check spelling (names, brands, places, products, etc.) & grammar * Intuit gaps in conversation * Fix punctuation & formatting * Format sentencing (remove filler words) * Apply other custom rules ### What it doesn't do [#what-it-doesnt-do] * It doesn't listen to audio or process it * It is the not the same as the ASR dictionary. Instead the AI Proofreader acts after the ASR engines have processed the audio and provided the text. * It doesn't do fact checking. ## Examples [#examples] A great example of this is the works of Professor Tolkien. We shared some of the captions lines (one produced by a human captioner and one produced by ASR) with the AI Proofreader. | Source | Text | | --------- | ------------------------------------------------------------------ | | Original | "dwarves live in cause of doom" | | Corrected | "dwarves live in Khazad-dûm" | | Source | Text | | --------- | ----------------------------------------------------------------------------------------------------------------- | | Original | "Boromiet took over when scanning off fails" | | Corrected | "Boromir took over when Gandalf fell" | ## LLM Engines [#llm-engines] The AI Proofreader is based on top of the best LLM engines available. It makes use of the extensive world knowledge inherent in the LLM training data and it also applies the specific AI Context of the event, as configured in the [AI Context Builder](/docs/ai-services/ai-context). Engines available: | Provider | Model | | --------- | ------------------------------------------------ | | OpenAI | `GPT 4o Mini` `GPT 4o` `GPT 5 Mini` `GPT 5 Nano` | | Anthropic | `Claude 4 Sonnet` `Claude 3.5 Haiku` | We use different LLM engines because each has its own strengths and weaknesses. These models offer each a different balance between speed and depth of thinking. However, we currently limit the choice of models to the ones available in the OpenAI and Anthropic APIs strictly because, in our testing, these are the only publicly available models that are able to correctly follow the rules we set for the AI Proofreader. If you think other LLM models should be available, please [contact us](https://line21.com/contact) and we'll be happy to consider it. ## Before or After [#before-or-after] The AI Proofreader can run before or after the captions are distributed to destinations (and visible by the audience). The toggle "Required for Distribution" controls this. We recommend you keep this toggle on for all events. Proofreader option ### When should I have it on? [#when-should-i-have-it-on] If using machine translation, livestreaming captions or translations. This is because the text needs to be delivered to the audience as its final form, without further corrections. ### When can I have it off? [#when-can-i-have-it-off] If providing captions via browser or overlay. The destinations are the only ones that can display further caption corrections. ## Rules & Custom Rules [#rules--custom-rules] Proofreader rules The AI Proofreader comes with a set of rules that are applied by default. These rules are: | Rule | Default | Description | | ---------------------------- | ------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Check Names and Proper Nouns | `on` | The Proofreader will check text against names provided in the context. | | Allow Factual Errors | `on` | This toggle prevents factual corrections. | | Custom Rules | | If you have preferences, such as preferring lowercase for certain words, or want the Proofreader to filter out specific words, then you can use conversational language in this field to explain what you want. | ## AI Context [#ai-context] The AI Proofreader makes use of the [AI Context](/docs/ai-services/ai-context) to understand the event and the content. This context can be fine-tuned any time on the duration of the event, and it is automatically referred to when the AI Proofreader is running. ## Latency [#latency] It depends on the underlying LLM engine used, but typically it is less than 1 second (\~400-500ms). --- # AI Translations (/docs/ai-services/ai-translations) Line 21 supports automatic translations in 99 unique languages across all translation engines. We use a variety of translation engines to make sure translations are accurate and cover as many languages as possible. ## Translation engines [#translation-engines] | Translation engine | Supported languages | Dictionaries | | --- | --- | --- | | Azure | 96 | ✓ | | DeepL | 35 | ✓ | | Google | 79 | ✓ | | Open AI | 99 | – | ## Supported languages [#supported-languages] Line 21 supports 99 unique languages across all translation engines. | Language | Code | Translation Engines | | --- | --- | --- | | Afrikaans | AF | Google, Azure, Open AI | | Albanian | SQ | Azure, Open AI | | Amharic | AM | Google, Azure, Open AI | | Arabic | AR | DeepL, Google, Azure, Open AI | | Armenian | HY | Google, Azure, Open AI | | Azerbaijani | AZ | Google, Azure, Open AI | | Bashkir | BA | Google, Azure, Open AI | | Basque | EU | Google, Azure, Open AI | | Bengali | BN | Google, Azure, Open AI | | Bosnian | BS | Google, Azure, Open AI | | Bulgarian | BG | DeepL, Google, Azure, Open AI | | Catalan | CA | Google, Azure, Open AI | | Chinese (Simplified) | ZH | DeepL, Google, Azure, Open AI | | Croatian | HR | Google, Azure, Open AI | | Czech | CS | DeepL, Google, Azure, Open AI | | Danish | DA | DeepL, Google, Azure, Open AI | | Dutch | NL | DeepL, Google, Azure, Open AI | | English | EN | DeepL, Google, Azure, Open AI | | Estonian | ET | DeepL, Google, Azure, Open AI | | Finnish | FI | DeepL, Google, Azure, Open AI | | French | FR | DeepL, Google, Azure, Open AI | | Galician | GL | Google, Azure, Open AI | | Georgian | KA | Google, Azure, Open AI | | German | DE | DeepL, Google, Azure, Open AI | | Greek | EL | DeepL, Google, Azure, Open AI | | Gujarati | GU | Google, Azure, Open AI | | Haitian Creole | HT | Google, Azure, Open AI | | Hausa | HA | Google, Azure, Open AI | | Hebrew | HE | Google, Azure, Open AI | | Hindi | HI | Google, Azure, Open AI | | Hungarian | HU | DeepL, Google, Azure, Open AI | | Icelandic | IS | Google, Azure, Open AI | | Igbo | IG | Google, Azure, Open AI | | Indonesian | ID | DeepL, Google, Azure, Open AI | | Irish | GA | Google, Azure, Open AI | | Italian | IT | DeepL, Google, Azure, Open AI | | Japanese | JA | DeepL, Google, Azure, Open AI | | Kannada | KN | Google, Azure, Open AI | | Kazakh | KK | Google, Azure, Open AI | | Khmer | KM | Google, Azure, Open AI | | Kinyarwanda | RW | Azure, Open AI | | Korean | KO | DeepL, Google, Azure, Open AI | | Kurdish | KU | Google, Azure, Open AI | | Kyrgyz | KY | Google, Azure, Open AI | | Lao | LO | Google, Azure, Open AI | | Latin | LA | Google, Open AI | | Latvian | LV | DeepL, Google, Azure, Open AI | | Lithuanian | LT | DeepL, Google, Azure, Open AI | | Macedonian | MK | Google, Azure, Open AI | | Malagasy | MG | Google, Azure, Open AI | | Malay | MS | Google, Azure, Open AI | | Malayalam | ML | Google, Azure, Open AI | | Maltese | MT | Google, Azure, Open AI | | Maori | MI | Google, Azure, Open AI | | Marathi | MR | Google, Azure, Open AI | | Myanmar (Burmese) | MY | Google, Azure, Open AI | | Nepali | NE | Google, Azure, Open AI | | Norwegian | NO | DeepL, Google, Open AI | | Odia (Oriya) | OR | Google, Azure, Open AI | | Pashto | PS | Google, Azure, Open AI | | Persian | FA | Google, Azure, Open AI | | Polish | PL | DeepL, Google, Azure, Open AI | | Portuguese | PT | DeepL, Google, Azure, Open AI | | Punjabi | PA | Google, Azure, Open AI | | Romanian | RO | DeepL, Google, Azure, Open AI | | Russian | RU | DeepL, Google, Azure, Open AI | | Samoan | SM | Azure, Open AI | | Sesotho | ST | Azure, Open AI | | Shona | SN | Azure, Open AI | | Sindhi | SD | Azure, Open AI | | Sinhala (Sinhalese) | SI | Azure, Open AI | | Slovak | SK | DeepL, Google, Azure, Open AI | | Slovenian | SL | DeepL, Google, Azure, Open AI | | Somali | SO | Azure, Open AI | | Spanish | ES | DeepL, Google, Azure, Open AI | | Swahili | SW | Azure, Open AI | | Swedish | SV | DeepL, Google, Azure, Open AI | | Tagalog (Filipino) | TL | Google, Azure, Open AI | | Tamil | TA | Azure, Open AI | | Tatar | TT | Azure, Open AI | | Telugu | TE | Azure, Open AI | | Thai | TH | Azure, Open AI | | Turkish | TR | DeepL, Google, Azure, Open AI | | Turkmen | TK | Azure, Open AI | | Ukrainian | UK | DeepL, Google, Azure, Open AI | | Urdu | UR | Azure, Open AI | | Uyghur | UG | Azure, Open AI | | Uzbek | UZ | Azure, Open AI | | Vietnamese | VI | Azure, Open AI | | Welsh | CY | Google, Azure, Open AI | | Xhosa | XH | Azure, Open AI | | Yoruba | YO | Azure, Open AI | | Zulu | ZU | Google, Azure, Open AI | ## Adding a translation [#adding-a-translation] Translations can be added from the AI Services section of the project settings. Add translation ## Changing the translation engine [#changing-the-translation-engine] You can change the translation engine for a specific translation language. We automatically recommend one of the engines, but you can test and change it as you wish. If you want us to integrate a new translation engine that better supports your use case, please [contact us](https://line21.com/contact) and we'll be happy to consider it. ## Custom dictionaries [#custom-dictionaries] Most translation engines support custom dictionaries. You can use this feature to tweak the translation of any pair of languages. Useful for specific words of phrases that are not correctly translated by the machine translation engine, especially proper nouns (like brands, e.g. Shell). Dictionaries --- # AI Services (/docs/ai-services) For vendor retention, AI training commitments, EU data residency, and Zero Data Retention options, see [Data handling & retention](/docs/security/data-handling). Line 21 uses over 12 different AI models in the live captioning process. These are powering 6 different services: | Service |
Status
| Description | | ---------------------------------------------------- | -------------------------------------- | ------------------------------------------------------------------------- | | [Speech Recognition (ASR)](/docs/methods/asr) | `Active` | AI speech-to-text models that transcribe the audio to text in real-time. | | [AI Translations](/docs/ai-services/ai-translations) | `Active` | AI models that translate the text to different languages. | | [AI Context](/docs/ai-services/ai-context) | `Active` | AI models that generate context for the AI Proofreader and ASR Engines. | | [AI Proofreader](/docs/ai-services/ai-proofreader) | `Active` | AI models that proofread the text and make corrections. | | [AI Call Agent](/docs/ai-services/ai-call-agent) | `Active` | AI bot that joins a video call and transcribes the audio in real-time. | | [AI Dubbing](/docs/ai-services/ai-dubbing) | `Beta` | A speech-to-speech model that dubs the speaker into other languages live. | --- # Browser (/docs/destinations/browser) A browser page is the simplest, most versatile, way to display live captions. It has a number of advantages: * `instant delivery` - since it doesn't require any additional video encoding it's latency is extremely low * `multiple languages` supports by default - viewers decide which language they want to see * `no additional setup` required - the browser page is always available in all projects * `no extra cost` * `highly accessible` - viewers can use their own devices to follow the captions, or the organisers can show the page on a screen * `highly customizable` - you can style the colors, the fonts, the size, the alignment, all to your liking ## When to use browser [#when-to-use-browser] Browser page caption delivery works in a wide range of scenarios including: 1. in-person event where the organisers want to show the captions on a screen 2. in-person event where the organisers want to share the link with the audience 3. online event where the organisers want to share the link with the audience through chat As a general rule: if you do not need to display the captions in a video player, the browser page is the default destination. Other text-only destinations are [explored here](/docs/destinations/text-destinations). ## Supported languages [#supported-languages] The browser page can display [all languages supported by Line 21 as a source](/docs/languages/supported-languages). ## Setup [#setup] No setup is required. Every Line 21 captioning project has a browser page link automatically generated at no extra cost. Browser choose language On access, the browser page link requires two pieces of information: * the `project code` (which is usually a 6-character code already included in the shared link) * the desired `language` (which is typically not included in the link and that viewers need to select for themselves) ## Sharing the link [#sharing-the-link] Browser share link Three ways to share the link: 1. Copy the link from the dashboard. 2. Show a QR code. 3. Embed the page itself in another website or application. ## Settings [#settings] The browser page has two types of settings: * `general settings` available from the dashboard (that apply to all viewers) * `individual settings` available from the browser page (that apply only to the current viewer) When an individual setting challenges the general setting, the individual setting takes precedence. For example, if the general setting sets the background color to green but the individual setting prefers black background, the black background wins. ## General settings (dashboard) [#general-settings-dashboard] ### Arbitrary delay [#arbitrary-delay] In some niche cases, you may want to add an arbitrary delay to the captions. For example, if you need to fine-tune the synchronization of pre-recorded captions in a live environment. This is available from the dashboard. ### Preset styles [#preset-styles] Line 21 provides fixed preset styles with the most common colour combinations for users. You can choose one of the presets or customize your own. Browser preset styles Available presets (1): | Preset | Description | | ----------------- | --------------------------------------------------------------------- | | White / Black | White text on black background | | Yellow / Black | Yellow text on black background | | Black / Yellow | Black text on yellow background | | White / Blue | White text on blue background | | Yellow / Blue | Yellow text on blue background | | L21 Green / Black | Special Line 21 green text on black background (like in old monitors) | | Open Dyslexic | Dyslexic-friendly font with white text on black background | You can preview the styles in box (2). Change sample languages for further fine-tuning. ### Custom style [#custom-style] Browser captions What you can customize: #### Colors: [#colors] * font color * background color #### Font family: [#font-family] Choose from a wide range of fonts: 1. Google Fonts (visible to all your viewers) 2. Your system fonts (visible only to you) 3. Public Font URL (visible to all your viewers) #### Text: [#text] * font size * font weight * line height * text alignment #### Accessibility checker: [#accessibility-checker] Line 21 has a built-in accessibility checker which uses WCAG; this will help you avoid colour contrasts that will be difficult for the audience to read and keep you compliant. Typically you would aim for an AA or AAA contrast ratio. Anything below AA is considered problematic for screen legibility. ## Individual settings (browser) [#individual-settings-browser] The viewer can access the following from the top right of their browser page: Browser captions ### QR Code [#qr-code] An easy access link if the host/viewer wishes to share it with other attendees on their own devices nearby from this page. ### Connection Status [#connection-status] An indicator that the viewer network is connected to the Line 21 platform and is stable; if green then all is well. If red, then the viewer is not connected and should check their internet connection. ### Viewer Menu [#viewer-menu] The access point for viewer customisations. Browser options #### Language [#language] Viewers can freely switch between their language choices at any time. This will refresh the page and the next new line will be in the selected language. #### Fullscreen [#fullscreen] This is an important setting, especially if you are the host. Using this setting will remove the banner at the top of the page for viewers, and just show text. If you use a browser page F11 or fullscreen system option, it will fullscreen the page including the banner (which is not always wanted, though we think our logo looks quite nice…) #### Show previous captions [#show-previous-captions] The last lines of characters will be shown in the view history. You can configure how many previous captions to show (1–99). This cannot be enabled at the same time as [Auto-clear inactive captions](#auto-clear-inactive-captions). #### Auto-clear inactive captions [#auto-clear-inactive-captions] When enabled, the screen is automatically cleared if no new caption has arrived for a configurable amount of time (default: 3 minutes, range: 0.5–60 minutes). This is useful for events with natural pauses, ensuring the screen does not show stale text. This cannot be enabled at the same time as [Show previous captions](#show-previous-captions). #### Style customization [#style-customization] Similar to the dashboard options. Note: The WCAG accessibility checker is not visible to viewers here, as we assume any viewers making individual adjustments will know their taste and text legibility needs. --- # HLS streaming (/docs/destinations/hls-streaming) ## When to use HLS streaming [#when-to-use-hls-streaming] HLS streaming is the best solution if you want to integrate multiple caption tracks into your video stream and deliver them to your audience in real-time. It takes a RTMP or SRT ingestion and generates an HLS playlist that your audience can watch in a number of ways. Advantages: * Multiple caption tracks * Multiple audio tracks, with [AI Dubbing](/docs/ai-services/ai-dubbing) * Syncronize captions with the video stream * Embeddable in any webpage (our player or your own) * Accepts incoming RTMP or SRT streams * Scalable to any number of viewers due to CDN caching ## Supported languages [#supported-languages] The HLS streaming destination can display [all languages supported by Line 21 as a source](/docs/languages/supported-languages). ## Prerequisites [#prerequisites] * Set up a [RTMP or SRT ingestion](/docs/feeds/video) * Enable an [ASR engine](/docs/methods/asr) to generate captions (optional) ## Setup [#setup] ### Multiple languages [#multiple-languages] An HLS video stream can include one or **multiple caption tracks**. This allows you to provide: * captions for the video's original language * additional tracks for any translations you configure All tracks are accessible from the video player's settings menu as native captions (no side-loading). ### Multiple audio tracks [#multiple-audio-tracks] Alongside the original audio, an HLS stream can carry **dubbed audio tracks**, one for each language you dub into with [AI Dubbing](/docs/ai-services/ai-dubbing). Add them to the HLS destination under **Interpretation audio tracks**. Viewers then pick their language from the player's audio menu, and every dubbed track stays in sync with the video. ### Captions delay [#captions-delay] HLS Captions Delay The captions are usually synced with the video stream. However, should a drift occur, you can adjust the captions delay to synchronize them with the video stream. This can happen with ASR and especially with manually produced captions. Our interface allows you to **delay** or **advance** the caption stream to fine-tune synchronization. ## Access HLS stream [#access-hls-stream] After creating an HLS destination, we provide the video output in several formats. HLS Options ### HLS stream URL [#hls-stream-url] You receive the raw HLS playlist URL (**.m3u8**), which can be used with any compatible player such as: * [Video.js](https://videojs.org/) * [JW Player](https://jwplayer.github.io/jwplayer/) This option gives you maximum flexibility for custom integrations. ### Player embed [#player-embed] You can also use our built-in video player.\ From the HLS destination page, copy the **embed code** and paste the iframe into your webpage. ### Hosted player page [#hosted-player-page] HLS Player Page For the fastest setup, we provide a **hosted viewer page** containing: * the video player * optional chat Just share the URL with your audience — no integration required. --- # Destinations (/docs/destinations) Once captions are generated, they need to be sent to a destination for them to be displayed for the audience. Line 21 supports 5 categories of destinations: | Category | Text | Video | Languages | Description | | --------------------------------------------------------- | ------------------------------------- | ------------------------------------- | --------- | ------------------------------------------------------------------------------------------------ | | [Browser](/docs/destinations/browser) | ✓ | ✗ | All | A sharable browser page that displays captions. | | [Overlay](/docs/destinations/overlay) | ✓ | ✗ | All | A floating, borderless window that display captions on top of other content or applications. | | [Text Destinations](/docs/destinations/text-destinations) | ✓ | ✗ | All | API-based delivery of captions as text to various target platforms, systems, apps, pages | | [RTMP Streaming](/docs/destinations/rtmp-streaming) | ✓ | ✓ | Single | Encoding captions into a RTMP video and simulcast it to Facebook, Twitch, Twitter or YouTube. | | [HLS Streaming](/docs/destinations/hls-streaming) | ✓ | ✓ | All | Publish video and captions to an HLS video player, embeddable in your own website or standalone. | --- # Overlay (/docs/destinations/overlay) Overlay Destination The Overlay is a floating, borderless window with customizable colors and styling, designed to display captions on screen on top of other content or applications. This feature is part of the [Line 21 App](/docs/line-21-app), which you can install on your computer to manage your captions and other settings. Explore the [Overlay documentation](/docs/line-21-app/overlay) for more details. ## Supported languages [#supported-languages] The overlay can display [all languages supported by Line 21 as a source](/docs/languages/supported-languages). --- # RTMP streaming (/docs/destinations/rtmp-streaming) ## When to use RTMP streaming [#when-to-use-rtmp-streaming] RTMP Streaming is the most widely used protocol for live video distribution with captions. It is supported by all major social and video platforms (YouTube Live, Twitch, Facebook Live, X, etc.). Unlike a text destination, your original video will be **re-encoded** to include a single subtitle track that the destination platform can extract and display as closed captions. Line 21 provides a full cloud encoding service that handles this process for you. ## Prerequisites [#prerequisites] 1. Set up a [RTMP or SRT ingestion](/docs/feeds/video) feed. 2. Enable an [ASR engine](/docs/methods/asr) to generate captions (optional). ## 608/708 cloud encoding [#608708-cloud-encoding] Sending a live video feed to a social platform is done through an older captioning protocol called **CEA-608/708**.\ Line 21 currently supports a version of **CEA-608**, which comes with a few limitations: * Only **one caption track** is supported (this limitation is also imposed by major platforms such as YouTube Live and Twitch) * Encoding captions into the video introduces **additional latency**, making low-latency streaming incompatible * Mainly **Latin-based characters** are reliably supported ## Languages supported [#languages-supported] | CEA-608/708 Encoding Availability | | --- | | Albanian | | Catalan | | Czech | | Danish | | Dutch | | English (English (UK), English (US)) | | Finnish | | French (French (Canadian)) | | German | | Hungarian | | Indonesian | | Italian | | Kinyarwanda | | Norwegian | | Polish | | Portuguese (Portuguese (Brazil)) | | Romanian | | Sesotho | | Spanish (Spanish (Latin America)) | | Swedish | | Turkish | | Vietnamese | | Xhosa | ## RTMP destinations [#rtmp-destinations] We support most RTMP-based platforms including: Facebook, YouTube, Twitter/X, Twitch, Dailymotion, Vimeo, Kaltura, Edgio, Limelight, Brightcove, IBM, Microsoft Teams, Teradek Core, Dacast, Wowza Video, Akamai and custom RTMP destinations. Below you can find our main integrations. [Contact us](https://line-21.com/contact) if you require another RTMP-based destination. ### YouTube Live [#youtube-live] RTMP YouTube Live Captions #### Step 1: Configure YouTube Live to accept 608/708 captions. [#step-1-configure-youtube-live-to-accept-608708-captions] 1. Enable closed captions. 2. Set Caption sources to be "Embedded CEA-608/708" 3. Set Source language label. 4. Ensure "Normal latency" is selected. RTMP YouTube Line 21 #### Step 2: Configure the Line 21 project to send captions to YouTube Live. [#step-2-configure-the-line-21-project-to-send-captions-to-youtube-live] 1. Name the destination. 2. Select a video source (from the live video feeds created before, see Prerequisites). 3. Select the desired language for your captions (can be the original language or a translated language). 4. RTMP Url is pre-configured. 5. Add your YouTube Stream Key. ### Facebook Live [#facebook-live] Facebook Live allows you to create live video feeds for your audience.\ Visit the [Facebook Live documentation](https://www.facebook.com/help/1636872026560015/) for prerequisites and configuration details. Facebook Live automatically displays any caption track found in the RTMP stream, with no special setup required. Sometimes the captions are not displayed in preview, you will need to go live to see them. ### Twitch Live [#twitch-live] Twitch Live is a platform tailored specifically for live videos. Follow their instruction on how to [set up live streaming](https://dev.twitch.tv/docs/video-broadcast/). Setting up Twitch Live in Line 21 is similar to YouTube Live. ### X Live [#x-live] X (formerly Twitter) allows you to stream live videos directly to the platform.\ Follow their [documentation](https://help.x.com/en/using-x/x-live) to set up your first live stream. Setting up X Live in Line 21 is similar to YouTube Live. ### Vimeo Live [#vimeo-live] Vimeo Live is an alternative streaming platform similar to YouTube Live.\ To configure a live stream, follow their [setup guide](https://help.vimeo.com/hc/en-us/articles/12427289749137-How-can-I-create-a-live-stream). Setting up Vimeo Live in Line 21 is similar to YouTube Live. ### Custom RTMP endpoint [#custom-rtmp-endpoint] You can also use any custom RTMP server.\ Whether captions are displayed depends entirely on the receiving server’s capabilities. Setting up custom RTMP endpoint in Line 21 is similar to YouTube Live. --- # Text destinations (/docs/destinations/text-destinations) ## What are text destinations? [#what-are-text-destinations] Text Destinations Text destinations are targets (platforms, systems, apps, pages) that accept captions as **text** (and not video). The method of delivery is typically via an API credential. When such a destination is configured, be aware of the following aspects: 1. Timestamps: When allowed we distribute the captions alongside **timestamps** to allow the destination platform to synchronise them to the video. However, this is entirely dependent on the target's ability to process the captions in real-time. 2. Closed captions: The captions are typically visible to the audience as **closed captions** inside the event platform. Below you can find our main integrations. [Contact us](https://line-21.com/contact) if you require another API-based text integration. ## Supported languages [#supported-languages] In general, the text destinations can display [all languages supported by Line 21 as a source](/docs/languages/supported-languages). However, each target platform may have specific limitations on the languages it supports. Check with the platform's documentation for more details and definitely test before the event. ## Microsoft Skycap [#microsoft-skycap] Text Skycap Microsoft SkyCap is an internal tool used by Microsoft for particular communications needs. SkyCap allows for human or automated captioning. SkyCap's integration with Line 21 was specifically built for APAC languages, from 2023 onwards. Microsoft showrunners will need to provide a SkyCap ID for the 'box' they are using, and password. Microsoft showrunners will also need to ensure they have selected the correct matching language in SkyCap in order for captions to flow. ## YouTube API [#youtube-api] YouTube Live allows for 2 methods of providing closed captions for the audience: 1. **RTMP Encoded Captions**: this allows for an RTMP ingest stream containing already the captions encoded in 608/708 format. We cover this here. [RTMP streaming](/docs/destinations/rtmp-streaming#facebook-live). 2. **Caption Ingestion URL**: this allows for captions to be sent as time-stamped text. This method is covered below. ### Setup YouTube Studio [#setup-youtube-studio] You also need to setup the YouTube Studio to accept captions through this method. Text YouTube 1. Select "POST captions to URL" under "Captions source". 2. Choose the Source language (this is simply a label, Youtube will show whatever language we send based on the settings in Line 21 under the label selected here). 3. Copy the **Caption Ingestion URL**. You will need it later when you setup the Youtube destination in Line 21. 4. Ensure "Normal latency" is selected. See YouTube's own guide for more details: [Turning on Closed Captions for Events](https://support.google.com/youtube/answer/3068031?hl=en\&sjid=14861096646398475153-EU#zippy=%2Clive-caption-requirements-for-events%2Cturning-on-closed-captions-for-events) ### Setup Line 21 [#setup-line-21] In Line 21, when you choose the YouTube destination: Text YouTube 1. Give it the name you want the session to have. 2. Enter the **Caption Ingestion URL** or the **Stream Key** into the empty field. 3. Select the desired language for your captions. ### Restreaming (beta) [#restreaming-beta] In some niche cases, if you need to combine ASR from RTMP/SRT with text-based caption delivery to YouTube. In these cases you also need to handle the restreaming of the original video to the very same YouTube destination(s). We can enable this for you, [contact us](https://line-21.com/contact). Restreaming ### Notes [#notes] * It is important to have YouTube automated captions turned off, otherwise they will 'compete' with Line 21 captions and result in illegible captions * YouTube will only allow one language at a time via API token * If you are using a human captioner, you will need to have them join a 'source audiovisual call' to receive the real-time feed of the livestream. * If you are using automated captions through an RTMP / SRT stream, you should check the [RTMP streaming for YouTube](/docs/destinations/rtmp-streaming#youtube-live) documentation for more details. ## Zoom API [#zoom-api] Text Zoom Zoom allows third party captioning software to send captions in Meetings and Webinars. This can be human or automated captioning. ### Prerequisites [#prerequisites] * If you are using a human captioner from Line 21, they will need to join your call in order to hear the content to caption. * If you are using automated captions from Line 21, then our [AI Call Agent](/docs/ai-services/ai-call-agent) will need to join your call in order to process the audio-visual stream. Zoom call their API connection an **API Token** and this is retrieved from the Zoom session once the scheduled start commences. See [Zoom documentation](https://support.zoom.com/hc/en/article?id=zm_kb\&sysparm_article=KB0060368) for how to retrieve the API token. ### Setup [#setup] In Line 21, when you choose the Zoom destination: 1. Give it the name you want the session to have 2. Enter the API token into the empty field 3. Select the language your captions should appear in ### Notes [#notes-1] * Zoom Settings can only be accessed via browser to setup third-party captioning (not in the app) * It is important to have Zoom automated captions turned off, otherwise they will 'compete' with Line 21 captions and result in illegible captions * Zoom will only allow one language at a time via API token * Breakout rooms will require their own API token ## Microsoft Teams API [#microsoft-teams-api] Text MS Teams Microsoft Teams allows third party captioning software to send captions in their calls. In some cases, this service is locked to Enterprise accounts - you may need to check your account to access this feature. ### Prerequisites [#prerequisites-1] * If you are using a human captioner from Line 21, they will need to join your call in order to hear the content to caption. * If you are using automated captions from Line 21, then our [AI Call Agent](/docs/ai-services/ai-call-agent) will need to join your call in order to process the audio-visual stream. The AI Call Agent's support for MS Teams is under development. MS Teams call their API connection a **CART caption link** and this can be retrieved in advance of the session start. See [Microsoft documentation](https://support.microsoft.com/en-gb/office/use-cart-captions-in-a-microsoft-teams-meeting-human-generated-captions-2dd889e8-32a8-4582-98b8-6c96cf14eb47) for how to retrieve the CART caption link. ### Setup [#setup-1] In Line 21, when you choose the MS Teams destination: 1. Give it the name you want the session to have 2. Enter the API token into the empty field 3. Select the language your captions should appear in ### Notes [#notes-2] * It is important to have MS Teams automated captions turned off, otherwise they will 'compete' with Line 21 captions and result in illegible captions * MS Teams will only allow one language at a time via API token * Breakout rooms will require their own API token --- # Audio feeds (/docs/feeds/audio) ## When to use audio feeds [#when-to-use-audio-feeds] Audio feeds are important to ingest any audio source into Line 21 to let our [Automatic Speech Recognition models](/docs/methods/asr#asr-engines) to do their work. Whether you're an audio/video team connecting stage audio to Line 21, or you simply want to capture your computer microphone to transcribe a conversation, you're covered. The goal is to provide a clean audio feed where only voices are captured. In professional setups, you likely have access to dedicated microphones designed to capture only the speakers' voices. This audio is often routed into the computer through a professional audio mixer. The rule of thumb is: the cleaner the audio, the higher the accuracy. Avoid sending audience noise, ambient sounds or music together with speech whenever possible, as they will reduce [ASR accuracy and overall performance](/docs/methods/asr#troubleshooting). Even if you're not using professional equipment and are relying on your computer's microphone, try to keep the feed as clean as possible by isolating your voice from external noise. Line 21 offers multiple ways to ingest audio feeds. ## ASR Page [#asr-page] The ASR web page is a simple web page that allows you to ingest audio from any microphone available to your browser. It is a good way to quickly test the ASR models and get a feel for how they work. ASR Page The page is structured in 2 parts: * The top part is the monitor, it shows the audio being captured and distributed in real time. * The bottom part is the settings and controls section. ### Browser compatibility [#browser-compatibility] The ASR page is compatible with Chromium-based browsers: | Browser | Compatibility | | ------- | ------------- | | Chrome | ✅ | | Edge | ✅ | | Brave | ✅ | | Firefox | ❌ | | Safari | ❌ | ### Monitor [#monitor] The monitor shows the text being transcribed in real time as well as the history of the transcribed text as your audience sees it. ### Project Panel [#project-panel] This section helps you correctly identify the project by name and project code. As well as giving you access to 3 features: * A QR code to quickly access the [view page](/docs/destinations/browser) of the project * A chat button to open a chat with the Line 21 team (useful to coordinate with the team, ask questions, etc.) * A choice between "Discrete" and "Continuous" mode to display the captions in the monitor section above. ### Speech Recognition Panel [#speech-recognition-panel] Speech Recognition Panel This section allows you to handle the core functionality of the ASR page. 1. **Language selector**: select the language of the captions and the ASR model to use. 2. **Microphone selector**: select the microphone to use for the audio capture and visualize the audio signal passing through. 3. **ASR connection**: show the status of the ASR connection, the time elapsed since the connection was established and buttons to start and stop the ASR service. ### Service Health Panel [#service-health-panel] This section shows the health of the service and the status of the ASR connection. It tracks 3 metrics and computes an overall quality score: | Metric | Description | Possible Statuses | | ------------------- | ------------------------------------------------------ | ------------------------ | | Internet Connection | Status as reported by the browser | Online / Offline | | ASR Connection | Connection status to the ASR service | Connected / Disconnected | | WebSocket Latency | Latency of the WebSocket connection to the ASR service | Measured in milliseconds | You can also see a graph of the past 100 measurements of the overall quality score. ## Line 21 App [#line-21-app] The Line 21 App has a section dedicated to [audio ingestion](/docs/line-21-app/direct-audio). The advantage, compared to the browser, is that you have a dedicated app and monitoring it can be easier compared to a browser tab. --- # Video feeds (/docs/feeds/video) Line 21 provides all the tools you need to ingest and output video feeds. Whether you need to extract audio from a video conference or stream your content with captions on YouTube, you're covered. ## When to use video feeds [#when-to-use-video-feeds] Video feeds can be used for the following purposes: * Extract audio through [ASR](/docs/methods/asr) and create captions to deliver to your audience; * Add captions to your live video stream and re-publish them live on your favourite social platform (YouTube, Twitch,...) through [RTMP Streaming](/docs/destinations/rtmp-streaming); * Add captions to your live video stream and publish as [HLS video](/docs/destinations/hls-streaming) on your website; ## RTMP vs SRT [#rtmp-vs-srt] The video industry standards for real time video distribution is set on two main protocols: RTMP and SRT. `RTMP` (Real-Time Messaging Protocol) is the older protocol (dating back to the Adobe Flash era) but remains the most widely supported.\ All major streaming platforms, such as YouTube and Twitch, use it to receive live video streams and the same applies to many hardware encoders. Although some variants have attempted to reduce latency, RTMP generally suffers from a delay of **2–5 seconds** (and sometimes up to **15 seconds**).\ It only supports **H.264/AAC** codecs, and while it can technically include multiple caption tracks, most platforms support only **one** subtitle track. `SRT` (Secure Reliable Transport), on the other hand, is a newer technology designed for greater reliability and lower latency.\ Unlike RTMP, it's **codec-agnostic**. It focuses solely on transport and can carry any video or audio format.\ However, SRT adoption is still limited: while infrastructure providers such as **Cloudflare** or **Mux** support it, most social platforms still only accept RTMP.\ Even though SRT can technically carry multiple caption tracks, in practice many platforms ignore them, meaning support for captions can sometimes be even worse than with RTMP. | Feature | `RTMP` | `SRT` | | ------------------------ | ------------------------------------------------------ | ------------------------------------------------------- | | **Transport Layer** | TCP | UDP | | **Latency** | 2–5 seconds (can be higher) | As low as 100–500 ms (tunable) | | **Reliability** | Reliable via TCP, but high latency under poor networks | ARQ and FEC mechanisms for reliability with low latency | | **Security** | Basic (TLS) | Built-in AES-128/256 encryption | | **Codec Support** | H.264 + AAC | Any codec (container-agnostic) | | **Adoption / Ecosystem** | Legacy; still used by YouTube, Facebook, etc. | Modern; adopted by Cloudflare, Mux, OBS, and others | At Line 21, we handle the ingestion of your live stream in **any format** (SRT or RTMP) and take care of retransmitting it in the correct format to any destination.\ You don't need to worry about destination-specific differences: just send us your stream via RTMP or SRT, depending on your setup. ## Video Ingestion [#video-ingestion] Video Ingestion To enable video/audio ingestion, open the **Language Panel**, then click on the **ASR** `1` toggle to start processing. It will display your SRT `2` and RTMP `3` ingestion endpoints and secret. ## Recommended settings [#recommended-settings] Video size and resolution strongly depends on your final destination and we may limit some resolutions in some cases. | **Parameter** | **Recommended Value** | | --------------------------- | ------------------------------------------------------ | | **Resolution** | 360p to 1080p | | **Frame Rate (FPS)** | 25 fps – 30 fps | | **Video Codec** | H.264 (AVC) | | **Profile / Level** | High @ 4.1 | | **Rate Control** | CBR (Constant Bit Rate) | | **Video Bitrate** | 4 – 6 Mbps (1080p) • 2 – 3 Mbps (720p) • 1 Mbps (480p) | | **Audio Codec** | AAC-LC | | **Audio Bitrate** | 128 – 192 kbps | | **Keyframe Interval (GOP)** | 2 seconds (e.g. every 60 frames @ 30 fps) | | **Latency Mode** | Low Latency (for SRT) | | **Container Format** | FLV (for RTMP) / MPEG-TS (for SRT) | * **Network stability** is more important than raw bandwidth. Test your upload speed. * Avoid using **Wi-Fi** or **mobile data** for live streaming whenever possible. **NOTE**: If you only need to extract captions from the audio to distribute it to your audience in a webpage, consider sending only the audio track or use [direct audio ingestion](/docs/feeds). ## Video Destinations [#video-destinations] If you plan to distribute your video directly to users or through a social platform, use the following methods: * **RTMP encoding destination** for social platforms such as YouTube or Facebook.\ Supports **only one caption track**. * **HLS video player** for standalone playback on Line 21 or embedded on your own website.\ Supports **multiple caption tracks**. ### RTMP Destinations (CEA x08) [#rtmp-destinations-cea-x08] RTMP destinations is used when you need to add generated captions to your video and to send it to a social platform. Supported ddestinations include: * YouTube Live * Facebook Live * Twitch Live * X Live * Vimeo Live * Custom RTMP endpoint Read more about [RTMP Destinations](/docs/destinations/rtmp-streaming). ### HLS Destinations [#hls-destinations] An HLS destination creates a video playlist that your users can view through our video player.\ The player can be hosted on our pages or embedded in your website.\ We also provide a playlist URL that can be used with your own player of choice. This is the best solution if you're not constrained to existing platforms and want the highest level of customization. Unlike RTMP endpoints, HLS players support **any number of caption tracks**, making them ideal when you want to offer captions and translations within a single player. Read more about [HLS Destinations](/docs/destinations/hls-streaming). ## Monitoring [#monitoring] The [monitor page](/docs/project-management/monitor) will show you the status of your video streams as they pass through the Line 21 platform. You can check the bitrate, FPS, and other stream quality metrics. --- # Account (/docs/getting-started) Everything starts with opening an account. To do this, you simply need to click on the "Sign In" button in the top right corner of the screen. Line 21 does not separate between `sign up` and `log in` workflows. On a successful sign in, we will either create a new account for you or redirect you to the account you already have. OTP There are two methods to authenticate your account: * `OTP (One time password)`: Enter your email address and we will send you a code to your inbox. * `Google authentication`: Sign in with Google (if your email is @gmail.com or it is hosted on Google Workspace). ## OTP (One time password) [#otp-one-time-password] 1. Enter an individual email address - this becomes the **organisation owner**, so is best not on a shared inbox. You can add extra team members at no additional cost to an organisation. 2. Check inbox for OTP: you should receive an email with a 6 digit code to enter. This can be entered only once and changes with every login. 3. Enter the code and click on "Verify". ## Google authentication [#google-authentication] Simply hit the "Sign in with Google" button and follow the instructions. ## Video tutorial [#video-tutorial] [Video walkthrough](https://www.loom.com/share/ed947a3c81e74442a0e5d530f2961298) ## Notes [#notes] * there are no passwords to store or remember. We use OTP or Google authentication to keep the private information to a minimum. Nothing to hack or steal. * use the work email address if possible. * avoid the use of shared emails addresses (e.g. [info@yourcompany.com](mailto:info@yourcompany.com)) as they may not be associated with a single user. Individual traceability is important for security and compliance. --- # Organisations (/docs/getting-started/organisations) Once you have successfully [signed in](/docs/getting-started) to your account, you will be redirected to the initial page. Only now instead of seeing the `Sign In` button, you will see the `Dashboard` button. Clicking on the `Dashboard` button will open a menu to manage your organisations. Dashboard ## Creating an organisation [#creating-an-organisation] To create an organisation, click on the `Add Organisation`. This will open a modal where you can enter the name of the organisation and click on the `Add Organisation` button. Once the organisation is created, you will be redirected to the organisation's dashboard page. Notes: * Organisations names must be unique for the entire platform. * Avoid using generic names like "My Organisation", "My Company", "My Team", etc. Use instead the name of the company, a brand name or any other unique name. * You can have multiple organisations in your account. ## Managing organisations [#managing-organisations] To manage your organisations, click on the `Dashboard` button in the top right corner of the screen. This will open a dropdown menu where you can see all your organisations. Clicking on an organisation will take you to the organisation's dashboard page. ## Billing [#billing] **At a glance**: Registering creates a free account that allows you to test Line 21 in a limited trial mode. To upgrade to production mode, you need to provide your billing information and register a card. [Video walkthrough](https://www.loom.com/share/cdedc8923cfb43bcb14b3a4b05980a9d) The first time you choose `Billing` in your dashboard, you will be prompted to complete your billing setup by providing: **Billing Email**: This email can be the same as your account registration or another - your invoices will be directed here. **Company Name & Address**: Use the name of the legal entity (i.e. `Line 21` is the name of the organisation, but the legal entity is `L21 Limited`). **VAT Number**: Add the VAT number of the legal entity if available. This unlocks the Billing section of your dashboard, and gives you a trial account. ### Trial vs Active Modes [#trial-vs-active-modes] An organisation can be in `trial` or `active` mode.
| Feature | Trial mode | Active mode | | ------------- | ------------------------- | ----------------------- | | Functionality | Restricted (testing only) | Full access | | Card | - | Card & Billing required |
### Billing Details & Card [#billing-details--card] Add Your Billing Details Register your card and company billing details to unlock full account access for Line 21’s platform. Once clicked to register, the following box will pop up for your completion. Card Fill in the blank fields with the necessary information. Your card will be charged $0.00 to authorise future payments. Your account is now in `active` mode - **and will be billed on the 2nd of every month.** ## Team [#team] * You can invite other team members to your organisation. * Each organisation has one `owner` (the account that created the organisation), and all other members will be considered `members` with admin-level access. * You can transfer ownership of your organisation to another member at any time. [Video walkthrough](https://www.loom.com/share/51e5e7ff65464fd1942a63ee3afbc454) ## Usage [#usage] Our usage dashboard allows you to track your captioning minutes - how many, and of what type. * Input minutes (from direct captions) * ASR minutes * Translations minutes * AI proofreader minutes * RTMP Video encoding minutes * HLS Video player Minutes Notes: * This analysis is also available [per project](/docs/project-management/usage#project-usage), so you can see project breakdown of overall organisation usage. * Monthly usage is the basis for [billing](/docs/getting-started/organisations#billing). [Video walkthrough](https://www.loom.com/share/44e3991441834894ad371f5d969dca8e) --- # Languages (/docs/languages) Languages are the heart of Line 21's projects and the very purpose of providing live captioning services. [We support over 100 languages](/docs/languages/supported-languages), including major world languages and regional dialects. ## How we handle languages [#how-we-handle-languages] Line 21 handles languages in a very specific way, different from most other captioning services. Each language is a `channel` in the system. It receives captions from various sources (inputs): 1. From a human captioner 2. From a single or multi-language ASR engine 3. From a translation engine 4. From a meeting bot This ability to `combine different sources of captions` is what makes these language channels extremely powerful when it comes to delivering captions to selected destinations. > `Example`: let's say you have an event where both Korean and English will be spoken on the stage. You need to automatically generate captions in either of these languages, but also to cross-translate them into the other language. Finally, you need to display them separately for each language on two different screens. Line 21's software architecture is built to handle this scenario out of the box. In Line 21, when you add a language or a group of languages (aka. multi-language), you are essentially creating a new channel in the system with at least one source enabled (ASR or Human). All projects define at least one language input on project creation. ## Adding languages [#adding-languages] On project creation, you need to add at least one language input. Later you can modify the project to add more language inputs. There are two types of language inputs: * Single language input powered by human captioning * Single language input powered by ASR * Multi-language input powered by ASR. Adding languages ### Single language vs Multi-language [#single-language-vs-multi-language] Each standard project can define one or more language inputs. Each language input can be either single language or multi-language. | Input Type | Human Captioning | ASR | Description | | --------------------------------------------------------------- | ------------------------------------- | ------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ | | `Single Language` | ✓ | ✓ | It handles one input language, powered by human captioning or a [single-language ASR engine](/docs/methods/asr#single-language-asr). | | `Multi-language` | – | ✓ | It handles multiple input languages with a single [multi-language ASR engine](/docs/methods/asr#multi-language-asr) that can autodetect the languages. | Adding multiple ASR inputs In some cases, you might want to configure a project to support several single-language inputs instead of a multi-language input. > `Example 1`: a multi-session event with speakers of different languages and with human captioners switching between sessions. > `Example 2`: you have an event where you need to support several languages that are not, as a group, supported by a multi-language ASR engine, but there are models that support each of these languages individually. ### Translations [#translations] Another means to define additional languages supported by a project is to add [AI Translations](/docs/ai-services/ai-translations). Typically these are different from the input languages (e.g. the input languages are English and Spanish, but you want to add translations for French, German, Italian, etc.). However, there are important cases where you might want to add as translations the exact same languages as the input languages. > `Example`: you have an event where spoken languages are English and Spanish. You perform ASR for these two languages but your intention is to provide full English and Spanish captions for the audience. Which means it is not enough to caption, say, in English, just the parts where English is spoken, but you need to also translate the Spanish-language parts into English. Same goes for Spanish. The solution to this is to add the same languages (English and Spanish) as translations as well. ## Displaying languages [#displaying-languages] Once captions and created and translated in a specific language, they become available to be displayed by any desired [destination](/docs/destinations). Most of the destinations support any and all languages. See [Supported languages](/docs/languages/supported-languages). However, some destinations like RTMP streaming have a limited set of languages they support. See [RTMP streaming](/docs/destinations/rtmp-streaming#languages-supported). --- # Supported languages (/docs/languages/supported-languages) Line 21 has extensive support for 120 unique languages. We use UTF-8 encoding for all languages and a font family with excellent unicode coverage ([Inter](https://rsms.me/inter/)). This means that, technically, all languages are supported. However, the list of actually supported languages may be limited by availability of relevant services: * [human captioning](/docs/methods/human): not covering many languages * [ASR engines](/docs/methods/asr): they tend to cover a lot of languages, but not all dialects or accents * [translation engines](/docs/ai-services/ai-translations): good coverage, but here you should check if the pair source / target is supported * [AI dubbing](/docs/ai-services/ai-dubbing): 75 languages, through Gemini Live Translate * RTMP encoding: the most limited subset of languages (see [RTMP streaming](/docs/destinations/rtmp-streaming#languages-supported)) * [HLS streaming](/docs/destinations/hls-streaming): all covered * [browser](/docs/destinations/browser) and [overlay](/docs/line-21-app/overlay): all covered * [text destinations](/docs/destinations/text-destinations): in general, all covered, but you should check with the platform's documentation for more details and definitely test before the event | Code | Language | ASR | Translation | AI Dubbing | Human Captioning | | --- | --- | --- | --- | --- | --- | | AF | Afrikaans | ✓ | ✓ | ✓ | No availability | | SQ | Albanian | ✓ | ✓ | ✓ | Limited | | AM | Amharic | ✓ | ✓ | ✓ | No availability | | AR | Arabic (Arabic (Algeria), Arabic (Egypt), Arabic (Iraq), Arabic (Iran), Arabic (Jordan), Arabic (Kuwait), Arabic (Lebanon), Arabic (Morocco), Arabic (Palestine), Arabic (Qatar), Arabic (Saudi Arabia), Arabic (Sudan), Arabic (Syria), Arabic (Chad), Arabic (Tunisia)) | ✓ | ✓ | ✓ | No availability | | AR_AE | Arabic (United Arab Emirates) | ✓ | – | – | No availability | | HY | Armenian | ✓ | ✓ | ✓ | Limited | | AS | Assamese | ✓ | – | – | No availability | | AZ | Azerbaijani | ✓ | ✓ | ✓ | No availability | | BA | Bashkir | ✓ | ✓ | – | Limited | | EU | Basque | ✓ | ✓ | ✓ | High | | BE | Belarusian | ✓ | – | ✓ | No availability | | BN | Bengali | ✓ | ✓ | ✓ | Limited | | BS | Bosnian | ✓ | ✓ | – | No availability | | BR | Breton | – | – | – | No availability | | BG | Bulgarian | ✓ | ✓ | ✓ | No availability | | CA | Catalan | ✓ | ✓ | ✓ | High | | CEB | Cebuano | – | – | – | Limited | | ZH | Chinese (Simplified) (Chinese (Traditional), Cantonese (Traditional)) | ✓ | ✓ | ✓ | High | | CO | Corsican | – | – | – | Limited | | HR | Croatian | ✓ | ✓ | ✓ | No availability | | CS | Czech | ✓ | ✓ | ✓ | No availability | | DA | Danish | ✓ | ✓ | ✓ | High | | NL | Dutch | ✓ | ✓ | ✓ | High | | EN | English (English (UK), English (US)) | ✓ | ✓ | ✓ | High | | EO | Esperanto | ✓ | – | – | No availability | | ET | Estonian | ✓ | ✓ | ✓ | Limited | | FO | Faroese | ✓ | – | – | No availability | | FI | Finnish | ✓ | ✓ | ✓ | No availability | | FR | French (French (Canadian)) | ✓ | ✓ | ✓ | High | | FY | Frisian | – | – | – | No availability | | GL | Galician | ✓ | ✓ | ✓ | No availability | | KA | Georgian | ✓ | ✓ | ✓ | No availability | | DE | German | ✓ | ✓ | ✓ | High | | EL | Greek | ✓ | ✓ | ✓ | No availability | | GU | Gujarati | ✓ | ✓ | ✓ | No availability | | HT | Haitian Creole | ✓ | ✓ | – | Limited | | HA | Hausa | ✓ | ✓ | ✓ | No availability | | HAW | Hawaiian | ✓ | – | – | No availability | | HE | Hebrew | ✓ | ✓ | ✓ | Limited | | HI | Hindi | ✓ | ✓ | ✓ | Limited | | HMN | Hmong | – | – | – | No availability | | HU | Hungarian | ✓ | ✓ | ✓ | Limited | | IS | Icelandic | ✓ | ✓ | ✓ | Limited | | IG | Igbo | – | ✓ | – | No availability | | ID | Indonesian | ✓ | ✓ | ✓ | No availability | | GA | Irish | – | ✓ | – | No availability | | IT | Italian | ✓ | ✓ | ✓ | High | | JA | Japanese | ✓ | ✓ | ✓ | High | | JV | Javanese | ✓ | – | ✓ | No availability | | KN | Kannada | ✓ | ✓ | ✓ | No availability | | KK | Kazakh | ✓ | ✓ | ✓ | No availability | | KM | Khmer | ✓ | ✓ | ✓ | No availability | | RW | Kinyarwanda | – | ✓ | ✓ | No availability | | KO | Korean | ✓ | ✓ | ✓ | High | | KU | Kurdish | – | ✓ | – | No availability | | KY | Kyrgyz | – | ✓ | – | No availability | | LO | Lao | ✓ | ✓ | ✓ | No availability | | LA | Latin | ✓ | ✓ | – | No availability | | LV | Latvian | ✓ | ✓ | ✓ | Limited | | LN | Lingala | ✓ | – | – | No availability | | LT | Lithuanian | ✓ | ✓ | ✓ | Limited | | LB | Luxembourgish | ✓ | – | – | Limited | | MK | Macedonian | ✓ | ✓ | ✓ | No availability | | MG | Malagasy | ✓ | ✓ | – | No availability | | MS | Malay | ✓ | ✓ | ✓ | No availability | | ML | Malayalam | ✓ | ✓ | ✓ | No availability | | MT | Maltese | ✓ | ✓ | – | Limited | | MI | Maori | ✓ | ✓ | – | No availability | | MR | Marathi | ✓ | ✓ | ✓ | No availability | | MN | Mongolian | ✓ | – | ✓ | No availability | | MY | Myanmar (Burmese) | ✓ | ✓ | ✓ | No availability | | NE | Nepali | ✓ | ✓ | ✓ | No availability | | NO | Norwegian | ✓ | ✓ | ✓ | High | | NY | Nyanja (Chichewa) | – | – | – | No availability | | NN | Nynorsk | ✓ | – | – | No availability | | OC | Occitan | ✓ | – | – | No availability | | OR | Odia (Oriya) | – | ✓ | – | No availability | | PS | Pashto | ✓ | ✓ | – | No availability | | FA | Persian | ✓ | ✓ | ✓ | No availability | | PL | Polish | ✓ | ✓ | ✓ | High | | PT | Portuguese (Portuguese (Brazil)) | ✓ | ✓ | ✓ | High | | PA | Punjabi | ✓ | ✓ | ✓ | No availability | | RO | Romanian | ✓ | ✓ | ✓ | Limited | | RU | Russian | ✓ | ✓ | ✓ | High | | SM | Samoan | – | ✓ | – | No availability | | SA | Sanskrit | ✓ | – | – | No availability | | GD | Scots Gaelic | – | – | – | Limited | | SR | Serbian | ✓ | – | ✓ | Limited | | ST | Sesotho | – | ✓ | – | No availability | | SN | Shona | ✓ | ✓ | – | No availability | | SD | Sindhi | ✓ | ✓ | ✓ | No availability | | SI | Sinhala (Sinhalese) | ✓ | ✓ | ✓ | No availability | | SK | Slovak | ✓ | ✓ | ✓ | Limited | | SL | Slovenian | ✓ | ✓ | ✓ | Limited | | SO | Somali | ✓ | ✓ | – | No availability | | ES | Spanish (Spanish (Latin America)) | ✓ | ✓ | ✓ | High | | SU | Sundanese | ✓ | – | ✓ | No availability | | SW | Swahili | ✓ | ✓ | ✓ | No availability | | SV | Swedish | ✓ | ✓ | ✓ | High | | TL | Tagalog (Filipino) | ✓ | ✓ | ✓ | No availability | | TG | Tajik | ✓ | – | – | No availability | | TAQ | Tamasheq | ✓ | – | – | No availability | | TA | Tamil | ✓ | ✓ | ✓ | No availability | | TT | Tatar | ✓ | ✓ | – | No availability | | TE | Telugu | ✓ | ✓ | ✓ | No availability | | TH | Thai | ✓ | ✓ | ✓ | High | | BO | Tibetan | ✓ | – | – | No availability | | TR | Turkish | ✓ | ✓ | ✓ | High | | TK | Turkmen | ✓ | ✓ | – | No availability | | UK | Ukrainian | ✓ | ✓ | ✓ | High | | UR | Urdu | ✓ | ✓ | ✓ | No availability | | UG | Uyghur | ✓ | ✓ | – | No availability | | UZ | Uzbek | ✓ | ✓ | ✓ | No availability | | VI | Vietnamese | ✓ | ✓ | ✓ | High | | CY | Welsh | ✓ | ✓ | – | No availability | | WO | Wolof | ✓ | – | – | No availability | | XH | Xhosa | – | ✓ | – | No availability | | YI | Yiddish | ✓ | – | – | No availability | | YO | Yoruba | ✓ | ✓ | – | No availability | | ZU | Zulu | ✓ | ✓ | ✓ | No availability | --- # CART connector (/docs/line-21-app/cart-connector) ## Who is it for? [#who-is-it-for] If you are a professional stenographer, you are likely using dedicated CART software.\ The Line 21 App allows you to connect these applications directly to Line 21 in order to produce real-time captions. While you can still use our [Writer web page](/docs/methods/human#writer-page) to create captions, the desktop application offers several advantages. For example, it continues working even when the app is not in focus. ## Project Settings [#project-settings] Open the app, enter your Project Code, and click on **Connector**.
CART Connector
Choose the language you want to caption in, and then select the protocol used by your CART software.\ We currently support **Eclipse** and **Catalyst**. **Catalyst support is still in beta**, so you may encounter some issues. If you do, please [report them](https://line-21.com/contact) so we can improve compatibility. ## TCP/IP Method [#tcpip-method] Both supported protocols use **TCP/IP** as the connection method. After starting the server in the Line 21 App, configure your CART software to send its output to the same TCP/IP port. Once the connection is established, the Line 21 App will display a status indicator confirming that your software is connected. --- # Direct audio (/docs/line-21-app/direct-audio)
Direct audio
## When to use Direct Audio [#when-to-use-direct-audio] The Direct Audio feature allows you to ingest audio directly from your computer's sound card. This is useful when you're working with a digital mixer or a professional external audio interface and want to send high-quality audio to Line 21 for Automatic Speech Recognition. ## App Audio vs. Browser Audio [#app-audio-vs-browser-audio] The Direct Audio feature in the Line 21 App works similarly to the [ASR web page](/docs/feeds#asr-page). However, the desktop app provides a more direct connection to your audio devices and is especially useful if you want to keep the audio pipeline isolated from your browser environment. --- # App Overview (/docs/line-21-app) ## When to use the app [#when-to-use-the-app] The Line 21 App is a desktop application available for all major operating systems.\ It allows you to: * Ingest audio from your computer (usually for [ASR](/docs/methods/asr) purposes) * Produce [manual captions](/docs/methods/human) through your preferred CART software * Display captions on screen through a customizable overlay ## Installation [#installation] Line 21 is available for **Windows, MacOS, and Linux** (Debian/Ubuntu or AppImage). **Windows certificate** The Windows version currently has a temporary certificate issue. As a result, the operating system may show a warning during download or installation. You can safely ignore this message and proceed. You can download the app directly from:\ 👉 **[line-21.com/resources](https://line-21.com/resources)** | OS | Download Link | | --------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | |
Windows
| v{DESKTOP_APP_VERSION} | |
MacOS
| v{DESKTOP_APP_VERSION} | |
Linux (Debian/Ubuntu)
| v{DESKTOP_APP_VERSION} | |
Linux (AppImage)
| v{DESKTOP_APP_VERSION} | ## Project Code [#project-code] Once the application is installed and launched, you must enter the **Project Code** to access its features. Every project has a unique code shown in the main project dashboard, as illustrated below: Project Code ## Main Features [#main-features] After entering the Project Code, the application provides access to three main features: * **[CART Connector](/docs/line-21-app/cart-connector)** — Connect your preferred CART software and stream real-time steno input. * **[Overlay](/docs/line-21-app/overlay)** — Display captions using a highly customizable on-screen overlay, suitable as a standalone display or as a video overlay. * **[Direct Audio](/docs/line-21-app/direct-audio)** — Ingest audio directly from your system's audio devices. --- # Overlay (/docs/line-21-app/overlay) Overlay The Overlay is a floating, borderless window with customizable colors and styling, designed to display captions on screen on top of other content or applications. ## When to use the Overlay [#when-to-use-the-overlay] Because the window can be made fully transparent it can be used in certain scenarios: * Burned-in captions for live events: the Line 21 overlay can be used to display captions on screen on top of other content or applications via video production software such as OBS or vMix. * In-venue display: as a standalone display or as a video overlay. ## Supported languages [#supported-languages] The overlay can display [all languages supported by Line 21 as a source](/docs/languages/supported-languages). ## Prerequisites [#prerequisites] * A Line 21 project with a [Project Code](/docs/project-types#project-code). * The [Line 21 App installed](/docs/line-21-app#installation) on your computer. * The Line 21 App connected to the server via the project code. * Choose the overlay section. ## Settings [#settings]
Overlay Settings
You can configure these settings either before opening the overlay, or after it is open. The changes are applied immediately without the need to reopen the overlay window. ### Style Settings [#style-settings] You can fully customize the appearance of the captions.\ You may choose from predefined themes or adjust each setting individually. Options include: * Font color * Font size * Font family * Line height * Background color This allows you to match the exact visual style required for your production. ### Line Length [#line-length] The **Line Length** setting determines the maximum number of characters per line.\ This is useful when you need precise control over text wrapping, especially when integrating the overlay into a video workflow. ### Transcript vs Subtitle mode [#transcript-vs-subtitle-mode] The overlay offers two display modes: | Mode | Description | | ------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `Transcript` | Text remains on screen indefinitely and scrolls only when necessary. | | `Subtitle` | Text appears and disappears according to the caption timing, similar to standard movie subtitles. This mode is ideal when burning captions directly into a video. | ### Start at Top [#start-at-top] By default, captions appear from the bottom of the overlay window.\ If you prefer them to appear from the top, enable **Start at Top**. In both cases, once the captions have filled the available space, the new lines will scroll from the bottom of the page. ## Open / Focus the Overlay [#open--focus-the-overlay] Once all the settings are to your satisfaction, you can open the overlay by clicking the "Open overlay" button. A new window will appear waiting for captions to be displayed. When the overlay is open, you can click the "Focus overlay" button to bring it to the front. Due to its transparent nature, it is easy to miss if it is not in the foreground. ### Window Size [#window-size] You can resize the overlay window to your desired size. This is the main method for controlling the number of lines and the width of the overlay window. For each project, the width and height of the overlay window are saved in your browser's local storage. This means that if you close the browser and open it again, the overlay window will be resized to the last saved size. ### Test Captions [#test-captions] Test Captions Once the overlay is open, you can enable **Test Captions** from the top-right corner (1).\ This feature is useful for previewing style changes in real time. ### Ghost Mode [#ghost-mode] Ghost mode is the most important feature of the overlay (2). It hides the background of the overlay window itself, making it fully transparent. The lines of text carry their own background color, so you can still see the text clearly, but only if the text exists on screen. When no text is present, the window itself is invisible. Compare the overlay without (top) and with (bottom) ghost mode enabled. Ghost Mode If you lose track of the invisible overlay, simply click **Focus Overlay** to bring it back into view. --- # Speech Recognition (ASR) (/docs/methods/asr) ## What is ASR? [#what-is-asr] Automatic Speech Recognition (ASR) is the process of converting spoken audio into written text. Modern ASR systems are powered by advanced AI models, which have significantly improved in quality and accuracy over the last few years, even in challenging environments such as noisy backgrounds or heavily accented speech. These models are actively improved and new versions are released every couple of months. However, ASR can still struggle with overlapping speakers, and accuracy may be lower than that of professional human captioners. Despite these limits, ASR remains highly accessible, fast and cost-effective, making it a strong option for many use cases. ## ASR Engines [#asr-engines] Line 21 offers several ASR engines and automatically recommends the best fit based on the language spoken in your audio. Each engine has unique capabilities such as speaker diarization, custom vocabulary support, or audio event detection (e.g., music or laughter). Because every use case is different, Line 21 lets you choose the engine that matches your needs and helps you get the most out of speech recognition. Here a list of engines with capabilities: | ASR Engine | Diarisation | Code Switching | Audio Events | Supported Languages | | --- | --- | --- | --- | --- | | Speechmatics | ✓ | – | ✓ | 55 | | Gladia | – | ✓ | – | 104 | | Deepgram Nova | – | – | – | 68 | | AWS Transcribe | – | ✓ | – | 43 | ## Languages [#languages] Not all languages are supported by ASR technology, but Line 21 currently supports 126, covering most global use cases. We also support limited **[code switching](#code-switching)** for scenarios where multiple languages appear in the same audio. ### Supported languages [#supported-languages] | Language | Code | ASR Engines | | --- | --- | --- | | Afrikaans | AF | AWS Transcribe, Gladia, MAI_TRANSCRIBE_2 | | Albanian | SQ | Gladia | | Amharic | AM | Gladia | | Arabic | AR | Speechmatics, Gladia, Deepgram Nova, AWS Transcribe, MAI_TRANSCRIBE_2 | | Arabic (United Arab Emirates) | AR_AE | Deepgram Nova | | Armenian | HY | Gladia, MAI_TRANSCRIBE_2 | | Assamese | AS | Gladia, MAI_TRANSCRIBE_2 | | Azerbaijani | AZ | Gladia, MAI_TRANSCRIBE_2 | | Bashkir | BA | Speechmatics, Gladia | | Basque | EU | Speechmatics, Gladia, AWS Transcribe | | Belarusian | BE | Speechmatics, Gladia, Deepgram Nova | | Bengali | BN | Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Bosnian | BS | Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Bulgarian | BG | Speechmatics, Deepgram Nova, Gladia, MAI_TRANSCRIBE_2 | | Catalan | CA | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Chinese (Simplified) | ZH | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Croatian | HR | Speechmatics, Gladia, Deepgram Nova, AWS Transcribe | | Czech | CS | Speechmatics, Gladia, Deepgram Nova, AWS Transcribe, MAI_TRANSCRIBE_2 | | Danish | DA | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Dutch | NL | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | English | EN | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Esperanto | EO | Speechmatics | | Estonian | ET | Speechmatics, Deepgram Nova, Gladia, MAI_TRANSCRIBE_2 | | Faroese | FO | Gladia | | Finnish | FI | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | French | FR | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Galician | GL | Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Georgian | KA | Gladia | | German | DE | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Greek | EL | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Gujarati | GU | Gladia, MAI_TRANSCRIBE_2 | | Haitian Creole | HT | Gladia | | Hausa | HA | Gladia | | Hawaiian | HAW | Gladia | | Hebrew | HE | Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Hindi | HI | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Hungarian | HU | Speechmatics, Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Icelandic | IS | Gladia, MAI_TRANSCRIBE_2 | | Indonesian | ID | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Italian | IT | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Japanese | JA | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Javanese | JV | Gladia | | Kannada | KN | Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Kazakh | KK | Gladia, MAI_TRANSCRIBE_2 | | Khmer | KM | Gladia | | Korean | KO | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Lao | LO | Gladia | | Latin | LA | Gladia | | Latvian | LV | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Lingala | LN | Gladia | | Lithuanian | LT | Speechmatics, Deepgram Nova, Gladia, MAI_TRANSCRIBE_2 | | Luxembourgish | LB | Gladia | | Macedonian | MK | Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Malagasy | MG | Gladia | | Malay | MS | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Malayalam | ML | Gladia, MAI_TRANSCRIBE_2 | | Maltese | MT | Gladia | | Maori | MI | Gladia | | Marathi | MR | Speechmatics, Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Mongolian | MN | Speechmatics, Gladia | | Myanmar (Burmese) | MY | Gladia | | Nepali | NE | Gladia, MAI_TRANSCRIBE_2 | | Norwegian | NO | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Nynorsk | NN | Gladia | | Occitan | OC | Gladia | | Pashto | PS | Gladia | | Persian | FA | AWS Transcribe, Gladia, MAI_TRANSCRIBE_2 | | Polish | PL | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Portuguese | PT | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Punjabi | PA | Gladia, MAI_TRANSCRIBE_2 | | Romanian | RO | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Russian | RU | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Sanskrit | SA | Gladia | | Serbian | SR | AWS Transcribe, Gladia, Deepgram Nova | | Shona | SN | Gladia | | Sindhi | SD | Gladia | | Sinhala (Sinhalese) | SI | Gladia | | Slovak | SK | Speechmatics, Deepgram Nova, Gladia, MAI_TRANSCRIBE_2 | | Slovenian | SL | Speechmatics, Gladia, MAI_TRANSCRIBE_2 | | Somali | SO | Gladia | | Spanish | ES | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Sundanese | SU | Gladia | | Swahili | SW | Speechmatics, Gladia, MAI_TRANSCRIBE_2 | | Swedish | SV | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Tagalog (Filipino) | TL | AWS Transcribe, Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Tajik | TG | Gladia | | Tamasheq | TAQ | Deepgram Nova | | Tamil | TA | Deepgram Nova, Speechmatics, Gladia, MAI_TRANSCRIBE_2 | | Tatar | TT | Gladia | | Telugu | TE | Gladia, Deepgram Nova, MAI_TRANSCRIBE_2 | | Thai | TH | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Tibetan | BO | Gladia | | Turkish | TR | Deepgram Nova, Speechmatics, Gladia, MAI_TRANSCRIBE_2 | | Turkmen | TK | Gladia | | Ukrainian | UK | Deepgram Nova, Speechmatics, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Urdu | UR | Gladia, MAI_TRANSCRIBE_2 | | Uyghur | UG | Speechmatics | | Uzbek | UZ | Gladia | | Vietnamese | VI | Speechmatics, Deepgram Nova, Gladia, AWS Transcribe, MAI_TRANSCRIBE_2 | | Welsh | CY | Speechmatics, Gladia | | Wolof | WO | Gladia | | Yiddish | YI | Gladia | | Yoruba | YO | Gladia | | Zulu | ZU | AWS Transcribe | ### Code switching [#code-switching] In some situations, multiple languages are spoken within the same conversation. **Code switching** allows an ASR engine to expect and recognise more than one language.\ When enabled, the engine can detect which language is being used at any moment. Line 21 can also cross-translate between the languages detected.\ To use this feature, create a **[multi-language input](/docs/languages#single-language-vs-multilingual-projects)** to enable automatic code switching. ## Diarisation [#diarisation] Diarisation is the process of separating audio by speaker.\ When an engine supports and enables diarisation, each speaker's lines are identified and separated. Captions will reflect speaker changes clearly, helping viewers follow who is talking at any moment. ## AV Feeds [#av-feeds] ASR can process either **direct audio inputs** or **audio extracted from video**.\ If you're using one of the video ingestion methods available in Line 21, the system will automatically handle the audio for ASR.\ See the documentation on **[Audio feeds](/docs/feeds/audio)** and **[Audio/Video feeds](/docs/feeds/video)** for more details. ## Context [#context] Line 21's recognition quality is enhanced by our **[AI context](/docs/ai-services/ai-context)** system.\ The context contains all known information about your event: names, terminology, keywords, technical concepts, and more.\ When the context is processed, Line 21 generates a custom vocabulary for the ASR engine, ensuring that specific terms are understood correctly and improving accuracy beyond standard ASR models. ## Troubleshooting [#troubleshooting] If speech recognition accuracy is lower than expected, review the common issues below.\ Most ASR problems fall into one of these categories. ### Audio Quality Issues [#audio-quality-issues] Poor audio is the main cause of ASR errors. **Symptoms:** * Words sound distorted, muffled, or clipped * Volume is too low or too high * Background noise overwhelms speech * Audio “drops” or interruptions **How to fix:** * Make sure your microphone or audio source is correctly placed and not too far from the speaker. * Avoid using built-in laptop microphones for important events. * Reduce background noise (fans, music, chatter, AC units). * If using AV feeds, verify the audio track being ingested is the cleanest available. * Check your audio format and ensure it matches the engine's requirements. * Check internet connection and ensure it is stable. ### Missing Vocabulary [#missing-vocabulary] If specific words are not recognised (names, cities, products, institutions), the issue is usually the **AI Context**. **How to fix:** * Verify that your **AI Context** includes all relevant names, technical terms, abbreviations, and jargon. * Add difficult words with phonetic hints if needed. * Re-upload updated context before your event starts. * See **[AI Context documentation](/docs/ai-services/ai-context)** for best practices. ### Speaker Overlapping [#speaker-overlapping] ASR engines still struggle when people talk simultaneously. **Symptoms:** * Words become jumbled * Parts of sentences disappear * Accuracy drops only during fast interactions or debates **How to fix:** * Encourage speakers to avoid overlapping speech. * If possible, normalise the audio so speakers have similar volume. * Enable **Diarisation** (if supported by the chosen engine) to help distinguish speakers. ### Wrong Language [#wrong-language] If parts of the transcript appear in the wrong language, check your language settings. **How to fix:** * Confirm you selected the correct language for the project. * If the content includes multiple languages, configure **Code Switching** and create a **multi-language input**. * See **[Languages](/docs/languages)** for configuration guidance. ### Engine Limitations [#engine-limitations] Different engines have different strengths. **How to fix:** * Try switching to another ASR engine optimised for your language or use case. * Review which engines support diarisation, vocabulary, or special audio events. * See **[ASR Engines](#asr-engines)** for recommendations. ### Still having issues? [#still-having-issues] If the problem persists, [contact us](https://line21.com/contact) and we'll be happy to help you troubleshoot. --- # Human Captioning (/docs/methods/human) ## What is human captioning? [#what-is-human-captioning] Human captioners are the gold standard of accessibility. They go by many names, with many acronyms around the world for professional standards: * CART (Communication Access Realtime Translation) * STTR (Speech To Text Reporter) * STTI (Speech To Text Interpreter) This isn't a gold standard as a marketing term; CART and STTRs are recognised by regulations in varying countries (ADA, Equality Act 2010, EAA) as the standard for accessibility and being compliant for reasonable adjustment requests from your audience. Human captioning is phenomenal for high stakes events, and where accuracy is critical - whether this be keynotes, conferences or news announcements. We also see human captioners used where confidentiality is king, or in offline venues - courthouses, and business roundtables. It is also very present in the educational and corporate sector in the day to day - if the language is particularly dense or expert specific, students and professionals will rely on human captioners as part of their journey. This aspect is often funded as part of government schemes (Access to Work, DAA). With all this demand, it is unfortunate that it always exceeds the number of captioners available - the key is to make requests early, and prioritise captioning rather than saving it as a last addition. [Contact us to check availability of human captioners with Line 21](https://line-21.com/contact). ## Writer vs Respeaker [#writer-vs-respeaker] Most qualified professional human captioners fall into one of two groups: ### Stenographers [#stenographers] `Stenographers` use a specialist keyboard, and train for approximately 2-3 years to build their speed and accuracy, allowing them to achieve a speed of up to `350 words-per-minute`. Usually starting in the courthouse as a court reporter, many will then transition to provide real-time support for d/Deaf users. Some will specialise in particular fields to develop their dictionary further - pharma, for instance. Stenographers pride themselves on their accuracy, which requires good preparation and context from organisers. They are bound by code of confidentiality as standard, and are typically registered with at least one governing body in their country of origin. Regionally, most stenographers are found in the US, UK and Australia. In Europe, there is a velotype keyboard, but not all European language providers use it. Considered to be the highest standard of accuracy, stenographers will be the most expensive service, and will often need to work in pairs if the content exceeds 60 minutes, or does not have adequate breaks built in. ### Respeakers [#respeakers] `Respeakers` are also known as **voice writers**. Increasingly popular, they can achieve similar word-per-minute rates of captioning as stenographers. They use voice software to 'respeak' the content they are hearing, often in shorthand, to accurately deliver a solid caption flow for the audience. As with stenographers, respeakers will build a dictionary for their clients to increase their accuracy in the industries they support. Respeakers will also have undertaken qualifications to serve as an STTR, and should be happy to provide these by request. Respeaking is the most common form of human captioning in regions that did not develop a stenographic machine, or in countries where accessibility is now becoming more important - and voice writing is considered an easier tool to learn to use. Commanding a high standard of accuracy that often matches stenographers, respeakers will still come at a higher cost depending on the language, and will need to work in pairs if content exceeds 60 minutes, or does not have adequate breaks built in. ### Other types [#other-types] Some providers choose to run their audio visual feed through ASR and have humans tidy the text. This is not recognised as a CART service by the relevant bodies. It can be a cost effective middle ground, but there is no regulation or recognised qualification for this action. Quality will vary immensely and is dependent on the ASR flow. ## Language support [#language-support] Language support depends on the availability of human captioners in the region. The client has the option to: 1. Provide their own human captioners, in which case any language can be supported. 2. Use Line 21's own human captioners, in which case the client can choose from the languages supported by Line 21. | Language | Human Captioning Availability | | --- | --- | | Basque | High | | Catalan | High | | Chinese (Simplified) (Chinese (Traditional), Cantonese (Traditional)) | High | | Danish | High | | Dutch | High | | English (English (UK), English (US)) | High | | French (French (Canadian)) | High | | German | High | | Italian | High | | Japanese | High | | Korean | High | | Norwegian | High | | Polish | High | | Portuguese (Portuguese (Brazil)) | High | | Russian | High | | Spanish (Spanish (Latin America)) | High | | Swedish | High | | Thai | High | | Turkish | High | | Ukrainian | High | | Vietnamese | High | | Albanian | Limited | | Armenian | Limited | | Bashkir | Limited | | Bengali | Limited | | Cebuano | Limited | | Corsican | Limited | | Estonian | Limited | | Haitian Creole | Limited | | Hebrew | Limited | | Hindi | Limited | | Hungarian | Limited | | Icelandic | Limited | | Latvian | Limited | | Lithuanian | Limited | | Luxembourgish | Limited | | Maltese | Limited | | Romanian | Limited | | Scots Gaelic | Limited | | Serbian | Limited | | Slovak | Limited | | Slovenian | Limited | Notes: * Human captioners have different workflows depending on regions. A team can be anything from 1-6 captioners depending on the language. * A baseline setup is two captioners per audio track if an event exceeds 60 minutes. * Some languages, such as Arabic, do not have human captioners as a profession. In these cases, most subtitling work is done in post-production. * We specialise in multilingual support and thrive in switching ‘on the fly’ languages live for our clients. To check prices, availability, book a human captioner or for any other question, please [contact us](https://line-21.com/contact). ## How captioners connect [#how-captioners-connect] As a human captioner, you can connect to Line 21 in one of three ways: | Method               | Type | Who is it for? | | ---------------------------------------------- | ------------------- | ------------------------------------------------------------------------------------ | | [Writer Page](/docs/methods/human#writer-page) | Text input | Stenographers (Stenokeys) and respeakers (Dragon Naturally Speaking with text input) | | [Speaker Page](/docs/methods/human#speaker) | Voice input | Respeakers that prefer to connect directly their microphone | | [Connector](/docs/methods/human#connector) | Desktop application | Stenographers that use Eclipse Advantage | ### Writer Page [#writer-page] Writer Page 1. **Input box** where you can direct StenoKeys to output text on mouse cursor. Use `enter` key to send a new line. Other rules for automatic line breaks [can be configured in the project settings](/docs/methods/human#human-captioning-rules). 2. **Connection status indicator**: green when connected, red when disconnected. Shows if the connection is established, i.e. Line 21 is ready to receive text. 3. **Language selector**: select the language of the captions. Allows for fast switching between languages. 4. **Simulation mode**: enable/disable [simulation mode](/docs/methods/human#simulation-mode). Useful for technical testing. 5. **Visual settings**: options to highlight differences between the current revision and the original one, and to animate the differences. Useful to visualise what the AI Proofreader is doing. 6. **Caption monitor**: shows the captions as they are being send through Line 21. If you see here a caption it means that the viewers are also seeing it. 7. **QR Code**: share this code with the viewers to allow them to view the captions in real time. It takes them to a different page, the ["Caption View Page"](/docs/destinations/browser). 8. **Visual modes**: options to switch between different visual modes to display the caption in this page. "Discrete" shows the captions stacked up, "Continuous" shows the captions flowing in a continuous stream. 9. **AI Proofreader Context**: shows the context of the AI Proofreader. Useful to understand what the AI Proofreader knows about the content and the event. 10. **Chat**: this is an internal chat to communicate with the other captioners and the project support team. Useful to coordinate handovers, see notifications from the client, etc. ### Speaker Page [#speaker-page] Speaker Page This page allows to connect directly your microphone to Line 21. It uses ASR models to transcribe your voice into text. 1. **Status indicator**: green when connected, red when disconnected. Shows if the connection is established, i.e. Line 21 is ready to receive audio. 2. **Language selector**: select the language of the captions. Allows for fast switching between languages. 3. **Simulation mode**: enable/disable [simulation mode](/docs/methods/human#simulation-mode). Useful for technical testing. 4. **ASR Models**: show a list of [ASR models](/docs/methods/asr#asr-engines) available for the selected language. Default ASR engine is preselected. You can test which ASR models is best for your voice. 5. **Start / Stop voice recognition** 6. **QR Code**: share this code with the viewers to allow them to view the captions in real time. It takes them to a different page, the ["Caption View Page"](/docs/destinations/browser). 7. **Visual modes**: options to switch between different visual modes to display the caption in this page. "Discrete" shows the captions stacked up, "Continuous" shows the captions flowing in a continuous stream. 8. **AI Proofreader Context**: shows the context of the AI Proofreader. Useful to understand what the AI Proofreader knows about the content and the event. 9. **Chat**: this is an internal chat to communicate with the other captioners and the project support team. Useful to coordinate handovers, see notifications from the client, etc. Notes: * Use `Chrome` / `Edge` or other `Chromium`-based browsers for best performance. Avoid `Firefox` due to a bug in the voice input. * On first page load, make sure to allow the browser to access your microphones. ### Connector [#connector] The CART Connector requires the installation of the Line 21 app. See more details on the [CART connector documentation](/docs/line-21-app/cart-connector). ## Human Captioning rules [#human-captioning-rules] To aid the human captioner, we have developed a set of rules to decide when a linebreak should be inserted automatically (thus allowing the captioner to keep writing without having to use the "enter" key). The rules are: Human Captioning rules 1. `New line / Enter`: new line is detected when the "enter" key is pressed. 2. `Punctuation`: new line is detected when a punctuation mark is detected. 3. `Length`: new line is detected when the line length exceeds a certain number of characters. The rules are configurable in the project settings under `Languages` section. ## Simulation mode [#simulation-mode] Simulation mode For technical testing, you can enable simulation mode. When enabled, the captions are autogenerated at a pace mimicking a human captioner. It has a number of options: * `Skip Destinations`: skip the external destinations (i.e. YouTube) and send the captions directly to the viewer page. * `Skip Proofreading`: skip the AI Proofreading service and send the original captions. * `Timestamps as text`: show timestamps as part of the caption text. Mainly for debugging purposes. * `Custom SRT file`: instead of using the AI generated captions, you can upload a timed captions file (.srt). Useful to test in other languages than English. Note that we use the orange color for simulated text to distinguish it from the original text. --- # Overview (/docs/methods) One of the fundamental choices when setting up live captioning projects with Line 21 is to decide which method of captioning you want to use. There are three main methods of captioning: 1. [Human Captioning](/docs/methods/human) - hire a human writer or a respeaker; 2. [Speech Recognition (ASR)](/docs/methods/asr) - AI models that transcribe the audio to text; 3. [Caption Playback](/docs/methods/playback) - use an existing subtitle file for technical testing or pre-recorded live streaming; Each method has its own advantages and disadvantages, and the choice of method will depend on your specific needs and requirements. [Contact us](https://line-21.com/contact) if you are unsure. | Method | Advantages | Disadvantages | | ------------------------ | --------------------------------------------------------------------- | --------------------------------------------------- | | Human Captioning | High accuracy, human-like captions, ability to handle complex content | More expensive, limited scalability, higher latency | | Speech Recognition (ASR) | Fast, scalable, ability to handle large volumes of content | Can have lower accuracy if not customised properly | | Caption Playback | Perfect accuracy, automated | Inflexible, fit only for niche situations | Traditionally human captioning was considered the only sound solution for high stakes live events. However, with the advent of AI models that can transcribe audio to text with higher accuracy, at a fraction of the cost, and the ability to [proofread in real-time](/docs/ai-services/ai-proofreader), many organisations are now considering using AI-powered captioning for their live events. Line 21 supports both of these methods. --- # Caption Playback (/docs/methods/playback) Caption playback is a method to play a pre-recorded subtitle file (.srt or .vtt) during a live event. It is useful for technical testing, or pre-recorded live streaming without the need to schedule it in a [playlist](/docs/project-types/playlist). ## Setup and configuration [#setup-and-configuration] Caption Playback 1. In the `Assets` page, under `Custom Caption Assets`, select `Upload File` and add your .SRT or .VTT file. 2. Press `Play` to see the page dedicated to playing the subtitle file. ## Going live [#going-live] Going live Control the playback simply by pressing the `Play`, `Pause` and `Restart` buttons. Scroll through the timeline to see the captions at different times. Your viewers will see the captions flowing in the [captions browser page](/docs/destinations/browser). ## Options [#options] Options There are a few options you can configure to customize the playback: * **Autoscroll** — When on, the page is autoscrolling to the current line. * **Highlight Current Line** — When on, the current line is highlighted and previous lines are slightly faded. * **Line Numbers** — When on, the line numbers are shown. * **Show Time** — When on, the timestamps are shown. * **Style** — The colors, font size, font spacing, presets, etc. ## Playlists vs Playback [#playlists-vs-playback] A [Playlist](/docs/project-types/playlist) is a type of project that uses a mix of live audio-visual feeds to provide live captions, as well as pre-recorded content. It can generate captions from live content, and also play a subtitle file. It used caption playback functionality to play the subtitle file. Our [Playback](/docs/methods/playback) feature is a more direct method used to upload third-party (or Line 21) subtitle files and play those captions for your viewers without scheduling them in a playlist. --- # Assets (/docs/project-management/assets) Assets The assets page allows you to manage transcripts and caption files in two different ways: 1. Download transcripts from your project sessions. 2. Upload caption files for playback. ## Transcripts [#transcripts] At any moment, you can download the full transcript for your project. This is a cumulative transcript of every text that was recorded at any point in time during the project's existence. If your project contains multiple sessions, you can filter by date and time to download the transcript for a specific session. * You can download your file in each language included in active machine translation in the event. * You can download your file as `.SRT`, `.VTT`, or `.TXT` format. ## Playback Assets [#playback-assets] You can also upload caption files for **playback**. This is useful if you already have a subtitle file you want to use for your event, or if you want to use a subtitle file from a different source. See our [Playback](/docs/methods/playback) documentation for more information. --- # Chat (/docs/project-management/chat) Line 21 includes an internal chat function available to all team members in a specific organisation and project. The chat is persistent and will follow you throughout the project. It is useful for quick communication between team members, or for coordination during the event. Chat Features: 1. monitor who is online at any given time 2. send and receive messages --- # Project management (/docs/project-management) If you are a Project Manager, AV Technician or accessibility specialist, then you may wish to supervise the event and make sure everything is working as expected. We offer a variety of tools to help you do this. It doesn't have to be a solitary task - you can invite additional [team members](/docs/getting-started/organisations#team) to help you out at no extra cost. ## Live Event Supervision [#live-event-supervision] Most of the critical live event supervision can be done from the [Monitor page](/docs/project-management/monitor). Includes features such as: * Overall data flow * ASR connections * Captions mini monitor * Video stream monitor Additionally you can always coordinate with the team using the internal [chat function](/docs/project-management/chat). ## Post-event Features [#post-event-features] * [track project usage](/docs/project-management/usage#project-usage) * [track organisation usage](/docs/project-management/usage#organisation-usage) * [track billing information](/docs/getting-started/organisations#billing) * [download transcripts](/docs/project-management/assets) * [upload assets for caption playback](/docs/project-management/assets) ## What can go wrong? [#what-can-go-wrong] A live event is an inherently complex and dynamic environment, involving many different servers and system interacting in real-time. Critical and non-critical issues can be prevented: | Issue | Prevention | | ---------------------------- | --------------------------------------------------------------------------------------------------------------- | | Poor audio quality | Ensure the audio source is clean and stable. This usually affects the quality of the ASR service. | | Poor internet stability | Ensure the internet connection is stable, test connectivity in advance and, if possible, during the live event. | | Human captioner connectivity | Test the captioner can successfully connect their equipment to the platform. | --- # Monitor (/docs/project-management/monitor) ## The System Monitor [#the-system-monitor] Service Monitor Our System Monitor gives you a real-time visual of our system flow. This is a direct representation of the data flow between the different services and destinations in your project, as per the configuration you have set up. From here you can see the status of all active inputs, AI processes, and outputs. ## ASR Connections [#asr-connections] ASR Connections Here you can track: * All active ASR connections - duration, connection source * Previous ASR connections (enable "Show inactive" toggle) This includes different users and sources, so for instance, if you have one user on a web connection and one via App, it will display here. You can also disconnect any running ASR session. ## Captions mini monitor [#captions-mini-monitor] Captions mini monitor The captioning mini monitor allows you to see the captions flow in real-time, without needing to have a view page open. This page will: * Show you real-time captions in the language of origin * Show you real-time translations in the secondary languages * Show you original text as `partial` in italics, before it becomes `final` after editing and is pushed * Monitor whether the [AI Proofreader](/docs/ai-services/ai-proofreader) is active and making corrections By clicking the arrow at the top right of this section, you can open a new page which is dedicated exclusively to revisions of all types. There, you can contrast the original text with the edited. Great for clients to see how their context held up and the original accuracy! ## Video Player monitor [#video-player-monitor] Video Player monitor When video destinations are enabled, you can monitor the video stream in real-time. You can see the video as it passes through our platform after any encoding was performed. If a video is present here, you are sure that we receive the [video feed](/docs/feeds/video) and that [video encoding](/docs/destinations/rtmp-streaming) is successful. This will also provide you with connection status information in the livestream, such as the bitrate of the audio and video streams. --- # Usage (/docs/project-management/usage) Line 21 is structured as a pay-as-you-go platform, meaning you only pay for what you use. We offer a variety of tools to help you track your usage. ## Project Usage [#project-usage] Project Usage Each project has its own usage dashboard, which allows you to track your usage for that particular project. We are tracking 6 different types of usage: * Input minutes (from direct captions) * ASR minutes * Translations minutes * AI proofreader minutes * Video encoding minutes (RTMP) * Video player minutes (HLS) ## Organisation Usage [#organisation-usage] Organisation Usage Similar to the project usage, the [organisation usage dashboard](/docs/getting-started/organisations#usage) allows you to track your usage for your entire organisation, across all your projects. The monthly billing information makes direct use of this data. Same metrics are tracked as in the project usage dashboard. --- # Project Types (/docs/project-types) When you create an organisation, your dashboard will look like this: Projects A project is a collection of settings and configurations that are used to create a live captioning session. You can use a 3-part mental model to understand how projects work: 1. `Inputs` — how captions are generated (methods, feeds, languages) 2. `Transformations` — how captions are transformed (AI services for proofreading, translations) 3. `Outputs` — where captions are delivered (destinations: Youtube, RTMP Encoding, Display in a browser, etc.) Notes: * Projects are usually not time-limited, they can be used immediately after creation. Billing is based on usage metrics liks minutes consumed. There is no start / stop button. * Projects can be archived if you don't need them anymore. * Projects should have unique names. (e.g. "Coachella 2025, Week 1"). ## Project Types [#project-types] To make it simpler for our users, we've made template projects of our most common use cases - making it easier than ever for you to set up for your events! - [Standard](/docs/project-types/standard): The default project type for most use cases, from simple to complex live captioning scenarios. - [Playlist](/docs/project-types/playlist): For scheduling a mix of caption playback and live captioning. - [Meeting](/docs/project-types/meeting): For online meetings that require better live captioning capabilities. ## Project code [#project-code] Every project has a unique Project Code that is used to identify the project in the Line 21 App. It is used in several places: * In any client-facing page that you may share with your audience, usually it is included in the URL. * In the [Line 21 App](/docs/line-21-app), when you open the project you want to use. You can copy the project code eithe from the Project List of your Organisation, or from the Project Details dashboard. Not sure which one to use? Email us on [hello@line-21.com](mailto:hello@line-21.com) to book a demo, or jump into our Discord for a video call. --- # Meeting (/docs/project-types/meeting) #### When would I use a Meeting Project? [#when-would-i-use-a-meeting-project] * Multilingual conference calls - a minimum of two ‘source’ languages * Multiple translation requirements #### Which destinations can I access in a Meeting Project? [#which-destinations-can-i-access-in-a-meeting-project] For our Meeting projects, you can access translated captions in our bespoke browser page. * Bespoke Browser * API Destination (1 Language) --- # Playlist (/docs/project-types/playlist) A playlist is a specialised project type for mixing pre-recorded and live content. It mantains all relevant functionality of a Standard Project. ## Use cases [#use-cases] * A venue showing live sports on a regular schedule, with pre-recorded spots for advertising or different content * Live tours or talks on a regular loop * Screenings with post-event Q\&A sessions Captions can be live, or submitted via `.srt` file for the Line 21 platform to run if you already have a file ready - great for VOD accessibility. A viewer can scan the playlist QR code and will be able to see when the next scheduled captions will be in the event, and which session is currently live. ## Setup [#setup] ### New project [#new-project] New Playlist Project Like a Standard Project, you will: * Name your project; * Choose the [Primary Language](/docs/languages); * Select `Playlist` as the project type. ### Settings [#settings] Main Settings 1. Edit the [Primary Language](/docs/languages), including whether it's [ASR](/docs/methods/asr), or [human](/docs/methods/human) captioning; 2. Select additional languages for [Machine Translation](/docs/ai-services/ai-translations), including adding custom dictionary; 3. Add context to the [AI Context Builder](/docs/ai-services/ai-context); 4. Use the [AI Proofreader](/docs/ai-services/ai-proofreader) and configure its rules during your captioning. Note: Follow the same steps as a Standard Project for connecting to ASR to feed [sound directly](/docs/feeds) or through [RTMP or SRT](/docs/feeds/video). ### Playlists [#playlists] Playlists 1. Add custom caption assets (.srt or .vtt files) to be scheduled in the playlists; 2. Add a playlist (name and timezone); 3. A viewer page link will be automatically generated for you, including the QR code. Share this with your guests / clients; 4. Add entries to the playlist. ### Playlist Entries [#playlist-entries] Playlist Entries 1. Title of the entry (visible to viewers); 2. Description (not visible to viewers, great for internal notes); 3. Type - whether `Live (RTMP / SRT Streaming`, or `VOD (video of demand)`; 4. Link custom caption assets (.srt or .vtt files) to the entry (if VOD entry type); 5. Add `Start Time` and `End Time` - this will appear on the Playlist viewer link, and control captions duration; Repeat the process for each entry you want to add to the playlist. Note: `SRT` means two different things - one is the file extension (e.g. `my_captions.srt`), the other is the protocol for streaming (`SRT - Secure Reliable Transport`). They are unrelated. ## Playlists vs Playback [#playlists-vs-playback] A [Playlist](/docs/project-types/playlist) is a type of project that uses a mix of live audio-visual feeds to provide live captions, as well as pre-recorded content. It can generate captions from live content, and also play a subtitle file. It used caption playback functionality to play the subtitle file. Our [Playback](/docs/methods/playback) feature is a more direct method used to upload third-party (or Line 21) subtitle files and play those captions for your viewers without scheduling them in a playlist. --- # Standard (/docs/project-types/standard) #### When would I use a Standard Project? [#when-would-i-use-a-standard-project] * Single track source language input to multiple destinations * Single track source language input to multiple translations * Multiple track language input to multiple destinations / translations #### Which destinations can I access in a Standard Project? [#which-destinations-can-i-access-in-a-standard-project] All destinations are available in your Standard Project type - and at the same time. More information on destinations and how they work can be found in our [destinations page](/docs/destinations). --- # Data handling & retention (/docs/security/data-handling) Customer knowledge-base article | Version 1.0 | June 2026 | Review cycle: every six months, or when a subprocessor changes ## Quick answers [#quick-answers] | Question | Answer | | ------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------- | | Is client content used to train third-party AI models? | No. Client content processed through the configured Line 21 pipeline is not used to train third-party AI models. | | Can specific vendors be excluded? | Yes. Vendor selection can be constrained per client engagement where policy requires it. | | Is EU data residency available? | Yes, depending on the selected speech-recognition and translation providers. | | Is Zero Data Retention (ZDR) available? | Yes. ZDR agreements are available for OpenAI and Anthropic engagements that require them. | | Are HLS recordings retained? | By default, HLS recordings are retained for 30 days on Line 21 infrastructure. This can be shortened or configured per client engagement. | This article explains how Line 21 handles client content in its live captioning, translation, and AI dubbing pipeline. It covers the main data categories processed, the third-party processors used, how long content may be retained, and whether submitted content is used for AI model training. This page focuses on the AI and media-processing pipeline. Standard account, billing, and commercial records are handled separately and are not shared with AI vendors. **Our position:** client content processed through the Line 21 pipeline is not used to train any third-party AI model. Vendors either commit to this in their standard terms or provide an opt-out that Line 21 has applied. ## Data categories processed [#data-categories-processed] * **Live audio/video streams** — the client's event feed, processed in real time for transcription and delivery. * **Transcripts, captions and translations** — text derived from the audio, passed through correction and translation services. * **Event context material** — glossaries, speaker names and domain terminology supplied by the client to improve accuracy; held on Line 21 infrastructure. * **Account and billing data** — standard commercial records, outside the scope of the AI pipeline and not shared with AI vendors. **Important distinction:** vendor retention refers to how long a third-party processor may temporarily retain submitted content. Line 21 retention refers to content stored on Line 21 infrastructure, such as HLS recordings or customer-supplied context material. These are governed separately and can be configured per client engagement. ## Status labels used in this review [#status-labels-used-in-this-review] | Status | Meaning | | -------------------------- | --------------------------------------------------------------------------- | | No training | The vendor does not use submitted customer content to train its models. | | Opt-out applied | The vendor allows an opt-out and Line 21 has enabled it. | | Under vendor clarification | Line 21 is confirming the vendor's exact AI-training position. | | Line 21 controlled | Data remains on Line 21 infrastructure and is not shared with an AI vendor. | ## Summary of processor commitments [#summary-of-processor-commitments] ### Speech recognition vendors [#speech-recognition-vendors] | Vendor | Content received | Content retention | AI training status | Notes | | -------------- | ---------------- | --------------------------------------------------------- | ------------------ | ------------------------------------------------------------------- | | Speechmatics | Audio for ASR | None | No training | Dashboard-level control. | | Deepgram | Audio for ASR | None | Opt-out applied | EU endpoint available for data residency requirements. | | Gladia | Audio for ASR | GDPR Standard Practice | No training | Compliance hub states customer data is not used for model training. | | AWS Transcribe | Audio for ASR | Auto deletion window of 90 days; earlier available by API | Opt-out applied | AWS service improvement use is disabled for Line 21 account. | ### LLM correction and proofreading vendors [#llm-correction-and-proofreading-vendors] | Vendor | Content received | Retention | AI training status | Notes | | ------------- | ---------------- | --------------------------------------- | ------------------ | ----------------------------------------------------------------------------------------------------------------------- | | OpenAI API | Text | Up to 30 days for abuse monitoring only | No training | Zero Data Retention agreement available for qualifying engagements | | Anthropic API | Text | 30 days | No training | Zero Data Retention agreement available for qualifying engagements. Anonymised usage metrics contain no client content. | ### Machine translation vendors [#machine-translation-vendors] | Vendor | Content received | Retention | AI training status | Notes | | ---------------------------- | -------------------- | --------- | ------------------ | ------------------------------------------------------------------ | | DeepL API Pro | Text for translation | | No training | GDPR-compliant; ISO 27001 and SOC 2 Type II certifications. | | Google Cloud Translation API | Text for translation | | No training | Advanced API offers regional endpoints. | | Azure AI Translator | Text for translation | | No training | GDPR-compliant; FedRAMP High certified. | | OpenAI API | Text for translation | | No training | Zero Data Retention agreement available for qualifying engagements | ### Recording capture and Line 21 storage [#recording-capture-and-line-21-storage] | Service | Content received | Retention | AI training status | Notes | | ---------------------- | -------------------------------------------------------------- | -------------------------------------------------------------------- | ----------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------ | | Attendee | Meeting audio/video and transcripts for online meeting capture | Audio/video retained for 5 days; transcripts retained until deletion | Under vendor clarification | Used only for specific online-meeting captioning workflows. Line 21 can exclude this workflow where client policy requires it. | | Line 21 HLS recordings | Video streams delivered to an HLS destination | 30 days by default on Line 21 infrastructure | Line 21 controlled, not used for training | Used for client playback and quality assurance. Retention and deletion-on-request are configurable per client engagement. | ## Client-configurable options [#client-configurable-options] Line 21 is a UK company and processes client data in accordance with UK GDPR and EU GDPR. The processing pipeline can be configured by vendor, region, and retention requirement. * **EU data residency** — EU endpoints are available for selected ASR and translation providers, including Deepgram EU and regionalised translation endpoints. * **Zero Data Retention** — ZDR agreements are available with LLM providers such as OpenAI and Anthropic for engagements processing sensitive content. * **Vendor exclusion** — Specific vendors can be excluded from a client's pipeline configuration where internal policy requires it. * **Custom retention** — HLS recordings and event context material can be shortened, deleted on request, or configured to meet client requirements. ### When Attendee is used [#when-attendee-is-used] Attendee is only used where Line 21 captions online meetings via a meeting bot. It is not part of every live-event workflow. Line 21 is clarifying Attendee's AI-training position directly with the vendor and evaluating customer-side storage options, under which recordings would reside on Line 21 infrastructure under Line 21's own retention policy. ## Review and contact [#review-and-contact] * **Last reviewed:** June 2026 * **Review cycle:** every six months, or whenever a subprocessor changes * **Owner:** Fabrizio Ruggeri, CTO and Co-Founder For questions about data handling on a specific engagement, [contact us](https://line-21.com/contact). *** ## Appendix: vendor reference notes [#appendix-vendor-reference-notes] These notes are intended for security, procurement, and legal reviewers who need the more detailed vendor-by-vendor position behind the public knowledge-base article. ### Speechmatics [#speechmatics] Realtime SaaS does not store audio, transcripts, or configuration data after processing completes. The workspace-level data-retention toggle is disabled in Line 21's account, so Line 21 traffic is not retained for service improvement. Line 21 does not route client content through Speechmatics' TTS preview product. ### Deepgram [#deepgram] Deepgram's default terms permit temporary storage of submitted audio for model improvement. Line 21 opts out programmatically via the `mip_opt_out` parameter on every request. With the flag set, audio is neither retained nor used for training. Deepgram offers a dedicated EU endpoint for data-residency requirements. ### Gladia [#gladia] Gladia commits contractually to never training on customer data and operates GDPR-compliant retention practices. No content-retention period beyond standard GDPR practice is published. ### AWS Transcribe [#aws-transcribe] Transcription jobs are automatically deleted after 90 days and can be deleted earlier via the Delete API. AWS's default permission to use content for service improvement is disabled for Line 21's account via the AWS Organizations AI-services opt-out policy. ### OpenAI API [#openai-api] API inputs and outputs are retained by OpenAI for up to 30 days solely for abuse monitoring, then deleted. Line 21 sets `store:false` on API calls, so content is not stored for later retrieval. Under standard API terms, submitted data is not used to train models. Zero Data Retention is available where required. ### Anthropic API [#anthropic-api] Standard retention is 30 days. Anthropic's standard API terms do not use customer data to train models. A Zero Data Retention agreement is available for engagements that require it. Anonymised usage metrics contain no client content. ### DeepL API Pro [#deepl-api-pro] Content is not written to persistent storage and is deleted immediately after translation. In exceptional error-debugging cases, content may be held encrypted for a maximum of 72 hours, then automatically deleted. API Pro texts are never used for model training. ### Google Cloud Translation API [#google-cloud-translation-api] Submitted text is not stored beyond what is required to serve the API response. Google states that content sent to the Cloud Translation API is not used to train or improve its translation models. The Advanced API offers regionalised endpoints for data-location control. ### Azure AI Translator [#azure-ai-translator] Azure AI Translator operates a no trace policy by default for API traffic. Customer content is not written to persistent storage and no record is kept in Microsoft data centres. Submitted text and audio are not used for training. ### Line 21 HLS recordings [#line-21-hls-recordings] Video streams delivered to an HLS destination are recorded and retained for 30 days on Line 21 infrastructure for client playback and quality assurance. This data is under Line 21's direct control, is not shared with AI vendors, and is not used for training. Retention and deletion-on-request are configurable per client engagement.