No Credit Card required
1000+
10,000+
40+
So natural the audience never clocks it's localized.
Lip sync that passes for the original
LipDub matches the mouth to the new language frame by frame, preserving the articulation, emotion, and pacing that make most dubs fall apart. 100% proprietary, and it holds quality from a 30-second clip to long-form.
The speaker's authentic facial expressions, preserved. Real results, not AI that looks like AI.
Broadcast quality,
under your control
Choose your model:
Pick your speed and quality
Choose a faster model for social and short-form, or a higher-fidelity model for premium work, and trade render speed for fidelity to suit the job. Produce in minutes, not weeks, and keep broadcast quality from a short clip to long-form.
The control that gets it past review.
Tune the expressivity of the mouth. Lip-sync only the regions you select, by clicking the timeline or entering a timecode. Reposition segments to re-align, and let the pacing indicator keep everything in sync. This is the difference between social-grade and broadcast-grade.
Works on any video, in any language
The world's best lip sync, built for AI video too.
Live-action, animated, or AI-generated: it all works. Already using avatars or AI video? Bring them in and lip sync them into every language at the same broadcast grade.
Already have the audio? Upload it.
Swap a line, fix a take, or replace the whole script without reshooting a single frame. Edit video dialogue as easily as text, and it feels like the first take, every time.
For Developers (API)
WALL OF LOVE
Campaigns that went global
Everything on this page, available via API
Translate, dub, and lip sync programmatically, at scale. Localization inside your own product or pipeline, for the platforms and teams that want to build in it.
Let's answer some FAQ's
Don’t hesitate to reach out if you have any questions
What is LipDub?
LipDub is a video localization platform. You upload a video you already shot (or generated), and we bring it into every market with the original person on camera, the original voice cloned to each language, and broadcast-quality lip sync. Marketing teams, agencies, online education platforms, and creators use LipDub to take what's already working in one market into every market they want to grow.
How does LipDub work?
Five steps. Upload the video you already shot (or generated). Translate the transcript into 80+ languages (refine any line in the editor before anything else generates). Clone the original speaker's voice, or pick one from the library. Lip sync the mouth to the new words. Export broadcast-ready versions for every market.
What kinds of content does LipDub work best for?
Long-form video where the person on camera matters: ads, courses, podcasts, panels, brand video, instructor-led training, talking-head content. Simple production (talking head, two-person conversation, podcast-style) with broadcast-grade source footage where the speaker is clearly visible.
How many languages does LipDub support?
80+ languages out of the box, with native pacing. Beyond that, LipDub is language-agnostic: if you upload your own audio, we can sync it to human lips. That includes minority languages, dialects, accents, and fictional languages. We've even seen users lip sync singing choir voices!
Is there a free trial?
Yes, with no credit card. The free tier includes 1 minute of content with AI dubbing, voice cloning, and lip sync. Paid plans are credit-based; see the pricing page.
What is AI lip sync?
AI lip sync re-animates a speaker's mouth, frame by frame, to match new audio, so a video translated into another language looks like it was shot in that language. LipDub preserves the speaker's articulation, emotion, and facial expression while it does it.
Is LipDub built on the same model as other lip sync tools?
No. LipDub's lip sync model is 100% proprietary, built for articulation, emotion, and texture fidelity, and it holds quality at length.
Will it look fake or hit the uncanny valley?
No. Native pacing out of the box and preserved facial expression keep it broadcast-grade. Real results, not AI that looks like AI. Tested with native speakers, the result is typically indistinguishable from the original.
Can I upload my own audio instead of translating?
Yes. Swap a line, fix a take, or replace the entire script while keeping the original on-screen performance. Dialogue editing runs through the same editor. If you've already standardized on another AI audio tool or are still casting voice actors, you can upload those audio files + the original video and get perfectly lip synced outputs.
How much control do I have over the output?
Tune the expressivity of the lip sync, lip sync only the regions you select (by timeline or timecode), reposition segments to re-align, and re-render only what changed after a script edit. You also have full control of the quality and texture fidelity of the lip synced mouth through training model selection.
What's the difference between the models?
Turbo is a fast, zero-shot model with no training step. Flash is faster and lower cost, close to Premium for most content. Premium is the higher-quality default. Ultra is the highest tier, on Enterprise. Turbo and Flash are free on every plan; Premium and Ultra add a one-time training fee per speaker.
What video formats do you support?
MOV or MP4 files, using the H.264 codec or Apple ProRes (422, 422 HQ, 4444, or 4444 XQ). Resolutions from SD up to 4K, at standard frame rates (23.976, 24, 25, 29.97, or 30 fps), in the sRGB or Rec.709 colorspace. A few things aren't supported: variable frame rate (convert to a constant frame rate first), interlaced or anamorphic footage, HDR, and files with multiple video streams.
What's LipDub's maximum video length?
Up to 180 minutes per video. 30 GB per video on the platform, unlimited size via API. There is also no restriction on how many videos you can process concurrently.
How does voice cloning work? How much source audio do you need?
LipDub clones the original speaker's voice from approximately 10 seconds of source audio. The cloned voice carries the speaker's character into every target language. You can also choose from a library of stock voices if voice cloning isn't required for your use case.
Can I edit translations before generating the final video?
Yes. The translation editor lets you refine and preview any line of the translated script before audio or lip sync generation. Useful for brand-specific terminology, tone adjustments, and fixing transcription errors. You can also preview audio and check pronunciation before lip syncing your video.
Can LipDub localize avatar-source video?
Yes. If your source video was generated by an AI avatar tool (HeyGen, Synthesia, or elsewhere), LipDub can localize that footage into additional languages. We localize whatever video is there, real or synthetic.
Can I replace dialogue instead of translating it?
Yes, you can start a Dialogue Replacement project where you can either write text-to-speech or upload your own audio.
Can I control the tone or style of the translation?
Yes. A custom prompt lets you guide tone, formality, and word choice, so the translation matches your brand voice instead of reading like a literal translation.
Can I generate the voice with text-to-speech?
Yes. As well as cloning the original speaker or choosing from the voice library, you can generate the dubbed audio with text-to-speech.
