Lip Sync Studio animates speech onto faces using audio as the driving signal. It supports two fundamentally different workflows: in Portrait Image mode you supply a still portrait photograph plus an audio file and the studio generates a talking-head video from scratch; in Video mode you supply an existing video alongside a new audio track and the studio re-syncs the lip movements to match the new audio. Nine dedicated models are spread across both modes, covering a range of resolutions, motion quality levels, and output styles.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/anil-matcha/open-generative-ai/llms.txt
Use this file to discover all available pages before exploring further.
Mode Toggle
Switch between the two input modes using the toggle at the top of the studio:| Mode | Input | Output |
|---|---|---|
| Portrait Image | Still portrait image + audio file | Animated talking-head video |
| Video | Existing video + audio file | Lipsync video with re-synced lip movements |
Image-based Models
These models take a portrait image and an audio file and produce a talking video. A text prompt is accepted as an optional parameter to guide the motion style.| Model | Endpoint | Resolutions | Prompt |
|---|---|---|---|
| Infinite Talk | infinitetalk-image-to-video | 480p, 720p | Optional |
| Wan 2.2 Speech to Video | wan2.2-speech-to-video | 480p, 720p | Optional |
| LTX 2.3 Lipsync | ltx-2.3-lipsync | 480p, 720p, 1080p | Optional |
| LTX 2 19B Lipsync | ltx-2-19b-lipsync | 480p, 720p, 1080p | Optional |
Video-based Models
These models accept an existing video and an audio file, replacing the original speech with the new audio track by resynthesising the lip movements.| Model | Endpoint |
|---|---|
| Sync Lipsync | sync-lipsync |
| LatentSync | latentsync-video |
| Creatify Lipsync | creatify-lipsync |
| Veed Lipsync | veed-lipsync |
| Infinite Talk V2V | infinitetalk-video-to-video |
Infinite Talk V2V also accepts an optional resolution picker (480p, 720p) and an optional prompt, making it the most flexible of the video-based models.
How It Works
Select a mode
Use the Portrait Image / Video toggle at the top of the studio to choose whether you are starting from a still image or an existing video.
Upload your portrait or video
Click the image upload button (Portrait Image mode) or the video upload button (Video mode) and select your source file. The file is uploaded to the Muapi file store and its URL is passed directly to the model.
Upload your audio
Click the audio upload button and select the audio file you want to use as the speech source. Supported formats include common audio types (MP3, WAV, etc.).
Enter an optional prompt
For models that support it, you can type a prompt to influence motion style or expression. Leave the field blank to let the model infer the motion from the audio alone.
Select a model and resolution
Choose the model that best fits your use case from the model picker. If the selected model exposes a resolution option, choose your preferred output resolution (480p, 720p, or 1080p where available).
Generation History
Lip sync generations are saved separately from other studio outputs in a dedicatedlipsync_history key in localStorage. If a generation is still pending when you close or reload the page, it resumes automatically on the next page load — you will not lose in-progress jobs.