Video Translator Review: Dub a Clip Into 28 Languages
Pick Video Translator to publish one clip in many languages in each speaker's own voice; set the source language over music, and expect unsynced lips.
Spicy Index
Not indexed
Freedom Score
Not yet tested
- Price
- —
- Median time
- Collecting data
- Success rate
- Collecting data
Video Translator listens to a talking clip, translates the speech and records it again in another language, timed to the original lines. Send video_url and target_language (28 to choose from). Turn on clone_voice and each speaker is re-voiced in a voice modelled on their own. keep_background_audio decides whether the music and effects stay under the dub, speaker_count (1 to 4) helps on dialogue, and glossary fixes the translation of names and recurring phrases. The picture is returned frame for frame, with a subtitle file of the translated lines as a second asset.
- SpicyAPI Engineering
- Tested Oct 16, 2026
- Methodology v3.3 →
- We sell these models; scores ignore price →
The verdict
It is billed per started minute of the source clip, so 70 seconds costs the same as two minutes; options do not change the price. At the time of writing that is $0.30 per started minute; see the model page for the live rate. The clip can run from 1 second to 10 minutes.
Key facts
- Content ladder: one bold-dialogue run, translated as written; the full ladder has not been run (see below).
- Price: $0.30 per started minute of the source clip at the time of writing, whatever the options; see the model page for the live rate.
- Input:
video_url(1 second to 10 minutes),target_language, plus optionalsource_language,clone_voice,keep_background_audio,speaker_countandglossary. - Output: the same video with the new audio, and an SRT subtitle file.
- Limits: lips are not re-synced to the new language.
- Sources: the parameter list is on the Video Translator model page.
What we liked and what held it back
What we liked
- The speakers still sound like themselves. In our two-person test the woman and the man came back in Spanish with their own timbre, and each line started where the original line started.
- The glossary works. Both lines that contained our fixed phrase came out with the translation we asked for.
- Background audio is yours to choose. With it on, the cellar ambience stayed where it was; with it off, the pauses between lines fell to near silence.
- The picture is untouched. A frame comparison of the dubbed clip against the source gave an identical match, so nothing is re-rendered.
- Subtitles come free. The translated lines arrive as an SRT in the same job.
What held it back
- Lips do not follow the new language. Only the audio is replaced. For close-ups, run the result through Lip Sync afterwards.
- Automatic language detection can fail. A whispered French line over a soft guitar made the task fail with a refund; the same clip with
source_languageset to French went through. Set it whenever there is music, an accent or long pauses. - Short lines can slip on grammar. One female line came back in the masculine form in our test; use the glossary for names and fixed phrases and read the subtitle file.
- The target must differ from the spoken language.
- Room tone can remain. On clips where the original sound is turned off, check the pauses for leftover noise.
What we tested
We ran four real clips on 16 October 2026; the three that completed were each billed as one minute. A 4-second line translated into German; an 8-second English dialogue between two people dubbed into Spanish with cloned voices, two speakers and a glossary entry; and an 8-second whispered French line over guitar dubbed into English. The French clip failed once on automatic detection and passed when the source language was set; the background was removed as asked, and the English line came out clear. Listening back, we checked each result against a transcription, and the text matched the subtitle file every time.
Completed dubs took about 80 to 86 seconds per short clip. Four short clips is a small sample, so read it as "the flow works and these are the failure modes we met", not as a score. Longer clips, music-heavy clips and less common language pairs are untested.
How far it goes
We ran one bold-dialogue clip, a short line from a performer, translated into German with default settings. It was translated and voiced as written, with no softening and no refusal. That is one run at one level, not our standard content ladder, so we make no claim about stronger content. You are responsible for having the rights to the footage and the voices in it.
Alternatives
- Need the mouth to match the new language? Lip Sync redraws the mouth to a new audio track; use it after dubbing.
- Want to rework the soundtrack yourself? Vocal Isolator splits a recording into vocals and background.
- Need sound for a silent clip? Video Sound Effects generates ambience and effects from the picture.
How Video Translator earns its Spicy Index
Each dimension below links to its full workings: the table row every point came from and why points were lost.
- Public leaderboards 45% of quality
- Not yet tested
- Blind output test 35% of quality
- Not yet tested
- Arena quality 20% of quality
- Not yet tested
- Capability 10.5% weight
- Not yet tested
- Freedom Score 30% weight
- Not yet tested
No exact match on a tracked public leaderboard yet. How it works
Not in our blind output test yet. How it works
Counts once the model has 100+ blind votes and a 95% interval under 200 Elo. How it works
Not scored yet. How it works
Freedom evidence is untested or insufficient. The Index retains the original quality and capability score, is marked preliminary and still receives a rank. Missing evidence is not zero. How it works
What Video Translator can do on SpicyAPI
Straight from the live model specification on SpicyAPI. Some creators offer more in their own apps; this is what you can call through our API today.
| Feature | Video translation |
|---|---|
| Native audio | No |
| Seed control | No |
| Negative prompt | No |
How long it really takes
Measured on real jobs over the last 30 days, from request to finished result. Times include our own test runs.
- Median time
- Collecting data
- 90th percentile
- Collecting data
- Wait per second of video
- Collecting data
- Success rate
- Collecting data
A time appears once at least 10 completed jobs are in. Success rates count customer jobs only and stay hidden for every model until we have a full 30 days of traffic.
Models to consider instead of Video Translator
SpicyAPI
Lip Sync
Need the mouth to match the new language?
- Spicy Index
- 65.6
- Freedom Score
- Not enough evidence yet
- Price
- —
SpicyAPI
Vocal Isolator
Want to rework the soundtrack yourself?
- Spicy Index
- —
- Freedom Score
- —
- Price
- $0.0012/ audio s
What do you think?
Was this review useful?
Log in to vote.
Comments
Comments are checked automatically and by our team. We remove spam, offensive content, personal attacks and anything that names a real person without consent. Report a comment and we review it within one business day.
Tried this model? Share what worked. You need a free account to comment, so we can keep spam out.
Video Translator FAQ
Try Video Translator yourself
Set every parameter in the Playground and see the exact price before you run it.
More reviews
See the leaderboards →
Video Upscaler
Pick Video Upscaler to deliver a clip at a larger tier before publishing; generate at a size that renders fast, then upscale only the cut you keep.
Tested Sep 27, 2026
Read the review
Talking Avatar
Pick Talking Avatar for a fast, cheap way to make a still portrait speak your own audio; it's not built for full-body movement or long-form video.
- Spicy Index
- 68.5
- Freedom Score
- Not enough evidence yet
Tested Sep 27, 2026
Read the review
Lip Sync
Pick Lip Sync when you already have footage and just need a new line coming out of the same mouth; it won't fix a clip that isn't a steady, single-shot take.
- Spicy Index
- 65.6
- Freedom Score
- Not enough evidence yet
Tested Sep 27, 2026
Read the review