One sample, whole video
Each voice source is cloned exactly once per render, reused for every line, and deleted when the render finishes — consistent voice, no per-line cloning, no clones piling up on the account.
CatalogLatency Script: transcription + translation only. Render: ~10–60s per minute of video plus one-time clone creation (~a few seconds). Both async. Async true Rate Limit 100 req/s per API key. Max Input 2 GB per source video; voice sample ideally 30–60s of clean speech.
cn8/voice-clone-dubbing
Identical to the dubbing product — video in, reviewable time-coded translation, ducked mix,
Learn more about Voice Clone Dubbing Videoasync (job)Instant previewaudiovideodubbing
- 1Video
- 2Script
- 3Render
Dub a video
The dialogue is transcribed and translated into editable, time-coded lines first — nothing is voiced until you approve it.
Dialogue is transcribed and translated into this language; lines already in it keep the original voice.
0.15/s of video for the script · rendering later 0.5/s + text-to-speech at 4/spoken second. Totals appear once the duration is known.
Overview
Identical to the dubbing product — video in, reviewable time-coded translation, ducked mix, dubbed video + audio out — with one difference at render: instead of a catalog voice you provide a voice sample (an audio file, or a video whose speech is extracted), and every dub line is synthesized in a clone of that voice. The clone is created once per render job and reused across all lines. Both stages run as async jobs.
Key capabilities
Same review-first flow
The script stage is shared with the dubbing product: edit the translation before anything is voiced.
Original soundtrack preserved
The original ducks only under dub lines; music and ambience survive everywhere else.
When to use it
Creator localization
Dub a creator's videos into new languages in their own voice.
Corporate spokespeople
Keep the same recognizable voice across localized announcement videos.
Input & output
input
Script: video URL + target_language (optionally options.speakers — how many people talk — and options.context — names/terms that guide transcription and translation, as in Dubbing). Render: the reviewed segments plus a clone source — a blanket options.clone_from_audio / clone_from_video, or per speaker in options.voices (clone_from_audio, clone_from_video, or clone_from_speaker: true to clone the speaker's own voice from the video).
JSON bodyvideo URL + voice sample (audio/video via upload-media)
output
video_url (dubbed MP4) + audio_url + render stats: clones_created, and clone_failed per speaker when a source could not be cloned.
JSON (script)S3 MP4 + M4A (presigned GET URLs)
Specs
Schema
Request body
video_urlrequiredstringPublic URL of the source video.
target_languagerequiredstringLanguage the dialogue is translated into.
enum: en, tr
expected_durationnumberVideo length in seconds — sizes the credit hold.
Pricing
Script and render billed per second; synthesis (including cloning) billed as core-tts usage on your account.
| Service | Unit | Price |
|---|---|---|
| Dubbing Script | second | $0.0015/sec |
| Voice Clone Dubbing Render | second | $0.005/sec + core-tts |
Async jobs are billed only when they complete successfully.