For Course creators & corporate trainers

Turn one lesson recording into a course every learner can follow

Turn a lesson recording into a version every learner can use: captions burned in for anyone watching without sound or who is deaf or hard of hearing, a dubbed audio track or translated captions for learners who don't follow your base language, and a sharper picture for recordings shot years ago. Accessibility captions for a one-hour lesson cost 7,200 credits, $176.40 to $324 — far under the $4,800 to $6,600 a professional subtitling house charges for the same hour in one language. Your team still owns the review: a caption or a dub is a draft until someone who speaks the language and knows the material has checked it.

The workflow

1
Decide what each lesson actually needs
Before uploading anything, decide per lesson: does it need captions in the language it was recorded in (for accessibility — anyone deaf, hard of hearing, or watching without sound), a dubbed audio track in another language (for learners who need to hear it, not read it), or just translated captions over the original audio (cheaper, and enough when learners can read your base language but don't speak it). Mixing these up is the easiest way to overpay: a dub nobody asked for costs as much as the captions everyone needs.
2
Caption the original for accessibility
Upload the lesson exactly as recorded — MP4, MOV or WebM, up to 30 minutes per file, so split a 60–90 minute lecture capture into parts first. Set the spoken language and leave "Translate subtitles to" on "Keep original language". You get back the same lesson with captions burned into the picture, at 2 credits per second: a 10-minute lesson is 600 × 2 = 1,200 credits. This is the version required for anyone watching without sound or who can't hear it at all, independent of which other languages you localize into.
Open the Subtitler →
3
Translate the audio for learners who need to hear it
For a market that needs to hear the lesson, upload the same original file and pick one of 25 target languages, including regional variants such as Spanish for Spain or Latin America. You get back a new video with a translated, time-aligned audio track. 2 credits per second, so a 10-minute lesson is 600 × 2 = 1,200 credits per language; one job makes one language, so five languages means five jobs.
Open the Video Translator →
4
Or translate just the captions and skip the dub
When learners can read your base language's captions but don't need to hear it spoken, set "Translate subtitles to" on the Subtitler to their language instead of making a dub first. One job transcribes, translates and burns in captions in the target language over the original audio — 2 credits per second, the same 1,200 credits for a 10-minute lesson, but half of what dubbing and then captioning that dub costs for the same lesson.
Open the Subtitler →
5
Caption every dub you publish, too
If you dubbed a language in the previous step, run that file through the Subtitler again with its new language set as the spoken language, so captions match what's actually playing. The accessibility reasoning from step two applies to every version you publish, not just the original — a Spanish dub still needs Spanish captions for a deaf or hard-of-hearing Spanish-speaking learner.
Open the Subtitler →
6
Leave code, commands and software UI alone
Neither tool touches on-screen text, so code samples, terminal output, config files and software menus stay exactly as recorded — there's no need to re-shoot screen capture per language. Say so once in the course introduction ("the code on screen stays in English") instead of repeating it in every lesson, and check that a dub's narration never names a translated button or command that no longer matches what's on screen.
7
Run the QA pass before anything goes live
Work through the per-language checklist below before publishing. A translator who doesn't know the subject can get the words right and the lesson wrong — technical terms, product names and anything tied to a certification exam need a reviewer who knows the material, not just the language.
8
Upscale only the recordings that actually need it
Cohort recordings from a few years back — a 720p webcam lecture, an old screen-share — upscale to 4K as a last step, on whichever finished version (original, dub or captioned) you'll actually publish. 3 credits per second: a 10-minute lesson is 600 × 3 = 1,800 credits. Skip it for anything recorded cleanly in the last couple of years; most screen-captured lessons don't need it.
Open the 4K Upscaler →

What a course hour costs

Live credit rates as of October 2026: captioning and dubbing both cost 2 credits per second, 4K upscaling 3. A credit costs from $0.045 (Starter Pack: $9 ÷ 200) down to $0.0245 (Pro plan: $49 ÷ 2,000 a month). A full course hour is 3,600 seconds, so accessibility captions or one translated-caption language both cost 3,600 × 2 = 7,200 credits — $176.40 to $324. A full dub plus captioning that dub costs twice that, 14,400 credits, $352.80 to $648 per hour per language. Upscaling an hour to 4K is 3,600 × 3 = 10,800 credits, $264.60 to $486.

Example: a 1-hour onboarding module captioned for accessibility in English, then made available in Spanish and French as translated captions only — no dub, since the audience can read but doesn't need to hear it spoken: 7,200 + 7,200 + 7,200 = 21,600 credits, $529.20 to $972 total. That's more than one Pro month's 2,000-credit allowance, so in practice it draws from top-ups too: twelve $59 Power Packs (1,800 credits each) cover it for $708.

For comparison, eLearning Solutions Lab's 2026 rate guide lists $80–$110 per finished minute per language for a professionally subtitled course and $300–$400 per finished minute per language for a fully produced human voiceover — $4,800–$6,600 and $18,000–$24,000 for one hour in one language. The module above, captioned in three languages combined, costs less than a single professional hour in a single language.

Dubbing and captioning cost the same per second, so the real lever is whether a market needs a dub at all — one that's comfortable reading cuts the bill in half.

What burned-in captions do and don't satisfy

Two rules now push captions from "nice to have" toward a requirement for a meaningful share of course creators. The U.S. Department of Justice's Title II rule requires state and local government entities — which covers public universities, community colleges and state-run training programs — to meet WCAG 2.1 AA, including captions on prerecorded video; after an April 2026 extension, the deadline is April 26, 2027 for jurisdictions of 50,000 people or more and April 26, 2028 for smaller ones. The EU's European Accessibility Act, in force since June 28, 2025, reaches further: it covers e-learning sold to EU consumers directly, not just government bodies, and expects captions in each market's own language, not just the language the course was recorded in.

Burned-in captions satisfy the core ask in both rules — text, synchronized, visible. They don't satisfy everything either one asks for: the EAA specifically wants captions to be "user-controllable," which an open caption burned into the video is not — a learner can't resize it, move it or turn it off — and neither rule is satisfied by captions alone if your own institution's policy also calls for a separate transcript. The Subtitler here only returns an MP4 with captions burned in, no separate SRT or VTT file; if a transcript or a closed-caption file is the actual requirement, run the lesson's audio through OpenAI's Whisper, free and open source, for an editable SRT on your own machine.

Neither rule was written with this tool, or any AI captioning tool, specifically in mind. Check with whoever owns compliance for your institution or your course platform before calling a course "accessible," especially if the audience includes a public institution or an EU market.

QA pass checklist for every language

A caption or a dub is a draft until someone who speaks the language and knows the material has checked it — a capable translator can get the words right and the lesson wrong.

Technical terms and jargon are translated the way your field actually uses them, consistent with any glossary you already have.
Burned-in captions don't cover on-screen code, commands, config or UI, and the narration never names a button, menu or command that no longer matches what's on screen.
Names, numbers, dates, units and currency are intact and in the target language's own format.
Captions match the audio that's actually playing — reused English captions on a new dub, or a caption pass run on the wrong file, is the most common mix-up.
Any term tied to a certification exam or vendor terminology stays recognizable, so a learner can still map it back to the exam.
Quiz and assessment text in your LMS is translated on its own if it's graded — captioning or dubbing the video doesn't touch the quiz.
Timing still lands on the right slide or demo step, not just the right line of audio.
A subject-matter reviewer, not only a linguist, signs off on anything with domain-specific content.

What not to translate

Some things should stay exactly as recorded, even when the rest of the lesson changes language.

Code, terminal commands, config files, regex and SQL — translating breaks them, and neither tool touches on-screen text anyway.
Software UI, menu paths and keyboard shortcuts shown on screen; check that a dub's narration doesn't describe a button or label that no longer matches the untranslated picture.
Product, brand and certification names, and vendor-specific terminology a learner needs to recognize on an exam.
Variable names, API endpoints, file paths and sample data in technical courses.
Legal and compliance language — disclaimers, safety warnings, policy citations — needs sign-off from whoever owns that content for the target market, not just a translation.

Limits to know before you commit

Plan a course launch around these before you rely on the workflow:

Up to 30 minutes per file on the Video Translator, Subtitler and 4K Upscaler — split longer lecture captures into parts before uploading.
One target language per Video Translator job; the only setting is the language, with no script review or glossary.
The Subtitler burns captions into the MP4 only — no SRT or VTT file; for an editable transcript, use Whisper (free, open source) instead.
Dutch, Polish, Swedish, Filipino and Malay are available for dubbing in the Video Translator but not for captions in the Subtitler.
Neither tool translates on-screen text, slides, code or software UI — that stays on your editor and your course materials.

Questions

Do burned-in captions make our course ADA or EAA compliant?

Not by themselves. They satisfy the core "captions on prerecorded video" requirement in both the U.S.'s Title II rule (which binds state and local government entities, with deadlines of April 26, 2027 or 2028 depending on jurisdiction size) and the EU's European Accessibility Act (in force since June 28, 2025, for e-learning sold to EU consumers). Neither is satisfied by captions alone if your institution also requires a separate transcript, and the EAA specifically wants captions to be user-controllable — resizable, repositionable, toggleable — which burned-in captions aren't. Check with whoever owns compliance for your institution or platform.

Can we get translated captions without paying for a full dub?

Yes. Set "Translate subtitles to" on the Subtitler to the target language instead of running the Video Translator first — one job transcribes, translates and burns in captions over the original audio, at 2 credits per second instead of 4 (2 for a dub, plus 2 to caption that dub). Use it when learners can read the target language but don't need to hear it spoken.

How long can a single lesson recording be?

Up to 30 minutes per upload on the Video Translator, Subtitler and 4K Upscaler. A 60- or 90-minute lecture capture needs to be split into parts first; split at the end of a topic, not mid-sentence, so captions and a dub still line up with the picture.

Do we get an SRT, VTT, or other transcript file?

Not from the Subtitler — it delivers captions burned into the MP4, nothing separate. If a transcript file is the actual requirement, for a stricter accessibility policy or so students can search the text, run the lesson's audio through OpenAI's Whisper, free and open source, for an editable SRT on your own computer.

What shouldn't we translate?

Code, commands, config files, software UI and anything a learner needs to recognize on a vendor exam — neither tool touches on-screen text anyway, but check that a dub's narration doesn't describe a button or command in words that no longer match what's on screen. Quiz and assessment text in your LMS is separate from the video and needs its own translation pass if it's graded.

What does a test cost?

The 20 free credits at sign-up cover a 10-second caption or dub test (10 × 2 = 20) — enough to check voice quality or caption accuracy on a short clip, with no watermark. After that, a $9 Starter Pack buys 200 credits (100 seconds), and the Starter plan $19 for 600 credits a month, or 300 seconds — 5 minutes — of captions or dubbed audio. Failed jobs are refunded.