Lipsync v3: Sync Any Face to Any Voice

Re-render mouth movements on existing footage or a single still image — even on profile shots, over-the-shoulder angles, and partially obstructed faces. Audio in, synced video out.

Try Lipsync v3
Lipsync v3
Lipsync v3

What is Lipsync v3?

Lipsync v3 is an AI lip sync model that re-renders mouth movements to match new audio. It works in two modes: video-to-video (re-syncs footage in up to 4K) and image-to-video (makes a still image speak). V3 processes the full shot at once, holding sync on profile angles and obstructed faces where earlier models slip.

What You Can Create with Lipsync v3

  • Dub Footage in a New Language

    Re-sync dialogue to a translated voiceover — the mouth matches the new audio while the rest of the frame stays untouched.

  • Make a Photo Speak

    Upload a portrait or illustration and an audio file. Lipsync v3 animates the mouth for the full length of the clip.

  • Fix Out-of-Sync Dialogue

    Replace or re-align audio on existing footage without reshooting. Works on profile and three-quarter angles.

  • Create Talking-Head Content from Stills

    Turn a single headshot into a talking video for ads, e-learning, or social — no camera, no studio.

  • Sync Across Difficult Angles

    Over-the-shoulder, side profile, partially obstructed faces — angles that break other lipsync tools.

Frequently Asked Questions

AI lipsync uses machine learning to re-render a person’s mouth movements so they match a new audio track. The model analyzes the audio waveform and the facial geometry in each frame, then generates new mouth shapes that stay in sync throughout the clip. Lipsync v3 processes the entire shot in one pass, which lets it hold sync on non-frontal angles and partially obstructed faces.

Yes. In image-to-video mode, Lipsync v3 takes a single image — a photo, illustration, or animated frame — and an audio file, then generates a video where the face speaks along to the audio for the clip’s full duration. The output length is determined by the audio track.

Lipsync v3 supports input resolutions up to 4K. It accepts video files, images (JPEG, PNG, WebP), and audio (MP3, WAV, M4A, AAC, OGG). Output is MP4. In video-to-video mode, the output matches the input video’s resolution and aspect ratio.

Lipsync v3 processes the full shot at once instead of frame by frame, which means it maintains accurate sync on profile shots, over-the-shoulder angles, and faces partially covered by hands or objects. Earlier models like Lipsync v2 Pro tend to slip on these non-frontal and obstructed scenarios.

Still have questions? We're here to help.