Put new audio on a video or image
Upload a talking clip or a portrait, add new audio, a recording or text-to-speech, and the mouth is re-synced to the new words.
- Dubbing and localization
- Fixing or replacing a line
- Making a character sing or speak
DomoAI · Lip sync
DomoAI Lip Sync: Match Speech to Your Video
New dialogue, dubs and restyles, with mouths that match
Give a clip new lines, dub it into another language, or restyle a talking video without losing sync. DomoAI analyzes the audio and reshapes the mouth to match every sound. It works on real people, anime characters and mascots.
3ways to sync
≤ 60 sper clip
Anylanguage audio
Choose a method
3 ways to lip sync with DomoAI
The right tool depends on whether you’re adding new audio or keeping the original speech.
Keep speech in sync while restyling
Turning a talking video into anime or cartoon? Switch on Lip Sync in Video to Video so the styled mouth still matches the original audio.
- Anime versions of vlogs and talking heads
- Restyled interviews and explainers
- Music performances turned cartoon
Automate it with the API
Send an image or a video plus audio to POST /v1/video/talking-avatar. Clips run 1–60 s, with aspect ratio and callback options.
- Dubbing pipelines
- Personalized video at scale
- In-app avatar features
Specs
Lip sync at a glance
| Visual input | Front-facing image or video with a visible face; works on real, anime and stylized characters |
|---|---|
| Audio input | Upload MP3, WAV or M4A up to 80 MB, record in the browser, or use text-to-speech (6 emotions, 6 tones) |
| Languages | Upload audio in any language |
| Length | Up to 60 s (Pro plan); API 1–60 s |
| Restyle lip sync | A toggle in Video to Video that keeps the original speech in sync |
| Output | Up to 1080p; 4K with the Video Upscaler |
| Speed | About 60 s for a 5 s clip; 60 s clips can take 10–15 min at peak times |
| API aspect ratios | 16:9, 9:16, 1:1, 4:3, 3:4 |
Step by step
Speaking character guide →How to dub or re-voice a clip
- 01
Pick the clip
A steady shot with one face, front-on or 3/4, the mouth visible and well lit.
- 02
Prepare the new audio
Record, use text-to-speech, or bring a voice file. Voice only, no music, silence trimmed.
- 03
Match the length
Keep the new audio close to the clip’s length. Use the timing checker below.
- 04
Upload & generate
Add the video or image plus the audio in Talking Avatar and generate.
- 05
Check “p, b, m”
The lips should close fully on these sounds. Review a short test first.
- 06
Finish
Add music and subtitles in your editor, and upscale if needed.
Interactive
Dub timing checker
New audio that’s much longer or shorter than the clip looks rushed or frozen. Check the fit before you generate.
New audio
Speaking pace
Inputs that sync well
- One face, front-on or 3/4 angle
- Mouth fully visible and evenly lit
- Head and shoulders with some space around them
- Steady camera with little motion blur
- Clean voice-only audio with natural pauses
- Clear speech at a natural pace
Inputs that cause problems
- Side profiles or faces turning away
- Hands, mics or hair covering the mouth
- Several people talking in one shot
- Background music or noise under the voice
- Very fast speech or dense rap lyrics
- Low-resolution or heavily compressed video
Use cases
All use cases →What people use video lip sync for
Dubbing & localization
Re-voice your own clips in other languages for new audiences.
Fix or update a line
Change a price, a date or a call to action without reshooting.
Anime talking videos
Restyle a vlog or interview into anime while the speech stays in sync.
Singing characters
Make a character perform your track, using a clean vocal stem.
Troubleshooting
Troubleshooting guide →When the lips don’t match
| Problem | Likely cause | Fix |
|---|---|---|
| Sync starts late | Silence at the start of the audio | Trim leading silence before upload |
| Lips drift or blur | Music or noise under the voice | Use an isolated, voice-only track |
| Mouth doesn’t close on p, b, m | Fast or mumbled speech | Re-record a little slower and clearer |
| Face changes over a long clip | Identity drift in long generations | Split into shorter segments |
| Odd teeth or jaw | Mouth covered or extreme angle | Pick a front-on shot with the mouth visible |
| Styled mouth off after restyle | Lip Sync turned off in Video to Video | Turn on Lip Sync and regenerate |
Consent comes first
Lip sync changes what a person appears to say. Only re-voice yourself, characters you own, or people who have clearly agreed. Never use it to impersonate anyone, spread misinformation or put words in a real person’s mouth. DomoAI’s policies prohibit this, and accounts can be suspended.
Be transparent
- Label AI-dubbed or AI-voiced content where platforms require it
- Keep the original files and permissions on record
- Use only voices you have the rights to, including cloned ones
- For client work, get written approval of the final script
Can DomoAI lip sync an existing video to new audio?
Yes. Upload a talking video or a portrait with new audio, and DomoAI re-shapes the mouth to match the new speech. The API’s talking-avatar endpoint accepts an image or a video plus audio, for 1–60 seconds.
How do I keep lip sync when restyling a video into anime?
In Video to Video (Restyle), turn on the Lip Sync setting. It keeps the styled mouth in sync with the original speech or singing. It works best with a visible mouth, a front or 3/4 angle and good lighting.
What audio formats are supported?
MP3, WAV and M4A up to 80 MB. You can also record in the browser or use text-to-speech with 6 emotions and 6 voice tones. Audio can be in any language.
Can I dub my video into another language?
Yes. Create the translated voice track (recorded, text-to-speech or from a voice tool you have rights to), keep it close to the original clip length, then lip sync the clip to the new audio.
How long can a lip-synced clip be?
Up to 60 seconds (60-second clips on the Pro plan; the API accepts 1–60 s). Split longer videos into segments, which also reduces identity drift.
How long does it take?
About 60 seconds for a 5-second clip. A 60-second clip can take 10–15 minutes at peak times.
Does it work on anime and cartoon characters?
Yes. DomoAI’s lip sync handles both realistic faces and stylized characters, such as anime, mascots and illustrations, as long as the mouth is clearly visible.
Is it okay to lip sync someone else’s video?
Only with their clear permission. Never use lip sync to impersonate someone or to make a real person appear to say something they didn’t. Label AI-dubbed content where platforms require it.
New words, perfectly in sync.
Re-voice, dub or restyle your clips with lip sync. Start with free credits.