The simplest way to localize an e-learning video without voice actors is to dub the original recording with AI, so each language keeps the trainer's own voice, and ship captions in the same language. Use text to speech when nobody speaks on camera, and subtitles alone when learners follow the original audio. Check the translated script first.
Updated 13 Nov 2026. Limits and credit rates below are as of Nov 2026.
| Method | What the learner gets | Best for | Main risk |
|---|---|---|---|
| Human voice actors (what you are avoiding) | A studio recording per language | Flagship content where top voice quality matters most | Cost and scheduling per language and per edit |
| Text to speech narration | A synthetic voice reads a translated script | Screen recordings, slide decks, no face on screen | Sounds impersonal for a trainer people know |
| AI dubbing (stock voice) | The video re-voiced in another language by a library voice | Talking heads where the trainer's identity matters less | Voice does not match the person on screen |
| AI dubbing with the trainer's cloned voice | The same trainer, speaking the new language | Trainer-led micro-lessons, onboarding, leadership messages | Needs the trainer's consent |
| Subtitles only | Original audio plus translated text | Short content, strong English audience, tight budget | Reading while watching slows learners on dense screens |
| On-screen text and slide translation | Translated labels, callouts and slides | Product walkthroughs where the UI language matters | Does nothing for the audio track |
A product walkthrough often needs narration plus translated on-screen text. A trainer intro needs a dub plus captions.
| Your video | Start with | Add |
|---|---|---|
| Trainer on camera | Dub in the trainer's cloned voice | Captions |
| Screen recording with voice-over | Dub the voice-over, or write a script and use text to speech | Translated on-screen text |
| Slides with narration | Dub the narration | Translated slides |
| Several speakers, such as an interview | Dub with up to 8 speakers, each in their own cloned voice | Captions (add speaker names when you review them) |
| Podcast style audio lesson | Audio translation | Transcript |
Two honest notes. First, people talking over each other are dubbed as one line, so avoid overlapping speech in source recordings. Second, product names and acronyms are where a translation slips, so turn on the script check for any lesson that names your product.
How to do it, step by step
- Cut the module into lessons. Clooney dubs up to 5 minutes per video today (1 minute on Starter). A 20-minute module becomes at least four separate dubs (4 x 5 minutes, or 20 x 1 minute on Starter).
- Clean the source. One speaker at a time, little background music, no overlapping speech. A clean source gives a cleaner dub.
- Get consent in writing. If you clone a trainer's voice, they should agree to it, for which languages and for how long. Keep your own record too.
- Upload the video to the dub tool at clooney.studio and choose the target language. Pick the trainer's own voice, or a library voice. Turn on Check the script first if the content is regulated, so you can read and edit every translated line before the voice is made.
- Add captions. Download the SRT that comes with the dub (or the captioned version), check names, product terms and numbers, and fix them. For VTT, or to re-time the lines, use the AI Subtitle Generator. Signed-in accounts get 100 free credits a month, no card needed. The 100 free credits cover transcription and subtitles only, not dubbing.
- Download the files and name them by language code, for example
onboarding-01-es.mp4andonboarding-01-es.srt.
How do you deliver localized videos to an LMS?
Clooney does not export SCORM packages and does not connect to an LMS. What you get is plain media files: the dubbed video, a version with captions burned in, the audio track on its own, and an SRT file. That works with most workflows:
- Separate course per language. Upload
-esand-devideos into a language-specific copy of the course. Simple, and easy to audit. - One course, language switch. Many authoring tools let you replace the media per language variant. Use the same lesson IDs so reports line up.
- Separate audio tracks. Some players accept alternate audio. Download the audio file and check what your LMS or player supports before you commit to this route.
Check one lesson end to end in your LMS before you localize the whole library.
Do dubbed e-learning videos still need captions?
According to the W3C, WCAG 2.2 success criterion 1.2.2 (Captions, Prerecorded, Level A) requires captions for prerecorded audio in synchronized media. A dub replaces the audio, but learners who are deaf or hard of hearing still need captions. Your captions must be in the same language as the dubbed audio. Review them: names, product terms and numbers are where automatic transcripts slip.
Do you need consent to clone an instructor's voice?
A cloned voice is a person's voice. Before you clone one:
- Ask the instructor for written consent that names the languages and the uses.
- Say what happens if they leave the company. Agree it up front.
- Tell learners that the audio was dubbed with AI. A line in the course intro is enough.
- Keep your records. Clooney requires consent to clone a voice, and the safety page explains how the process works.
Where Clooney fits, and where it does not
Clooney (clooney.studio) is an AI video translation, dubbing and subtitles studio for creators and training teams, with 42 languages and dubbing in the speaker's own cloned voice or a library voice.
Good fit: micro-lessons of up to 5 minutes (as of Nov 2026; 1 minute on Starter) where a real trainer speaks on camera; product and onboarding videos; teams that want a self-serve tool with no sales call, 42 languages, and the trainer's own voice. Plans start at $9/month, with a 7-day trial (card required, cancel before day 8 and pay nothing, per the terms).
Cost, in plain numbers. As of Nov 2026, a dub is 1 credit per second of video, with a minimum of 10 credits per dub. So a 45-second lesson costs 45 credits. See pricing for credits per plan. Failed jobs are refunded automatically (pricing FAQ). The one exception is a file longer than your plan allows, which keeps its 10-credit hold.
Not a fit today:
- Modules longer than 5 minutes in one file. Cut them first.
- SCORM or LMS integration, team seats and single sign-on. None exist yet.
- Training that must be word-for-word certified by a translator. Use a human reviewer, with or without Clooney.
We have not included a timed example from a Clooney job in this post. When we do, it will show the real clip and the real credit cost.
When to pick another tool instead
- No one speaks on camera and you want full control of the script: use text to speech with a script you translated and reviewed. Clooney has a text to speech tool, but any narration tool will do for this case.
- You need a new presenter, not your existing trainer: an avatar video tool makes new videos with a synthetic presenter. That is a different job from localizing footage you already shot.
- Your modules are 20 to 60 minutes and you cannot cut them: a tool that accepts longer files will save you work today.
- Your audience reads the original language comfortably and you only need access support: subtitles alone are enough. Subtitles use transcription credits only, and the 100 free monthly credits cover about 20 minutes.
- Legal or regulated content with sign-off per word: use a translation agency, then voice the approved script.
Try it on one lesson
Take one trainer-led lesson under a minute, dub it into one language and check it with a native speaker. Try a 10-second dub with no account, or see plans and credits. The 10-second preview is watermarked and only a taste of the voice.
Frequently asked questions
Short answers to what readers ask next.
Yes. You can dub the original recording with AI so each language keeps the trainer's own voice, generate narration with text to speech, or add subtitles only. Which one fits depends on whether a person speaks on camera and how formal the content is.
It can be, but treat the translation like any other regulated content: have a fluent reviewer check the script before release. Clooney has a Check the script first option, which stops after translation so every line can be read and edited before the voice is made.
Usually yes. WCAG 2.2 success criterion 1.2.2 (Level A) asks for captions on prerecorded audio in synchronized media, to serve people who are deaf or hard of hearing. A dub changes the language of the audio, so the captions need to match it.
No. Clooney gives you the dubbed video, a captioned version, the audio on its own and an SRT file. You upload those into your authoring tool or LMS the same way you upload any other media file.
Clooney dubs up to 5 minutes per video today, and up to 1 minute on the Starter plan. Longer modules need to be cut into lessons first.
Sources
- W3C: Understanding SC 1.2.2 Captions (Prerecorded), WCAG 2.2 · checked
- Clooney: Pricing · checked
- Clooney: AI Subtitle Generator · checked
- Clooney: Safety and consent · checked
- Clooney: Terms of service · checked