/ Models

← Media models

MuseTalk 1.5

Real-time lip sync — single-step latent inpainting rather than a diffusion sampler, so it is fast and small. Capped at 256×256, and the authors note some jitter and lip-colour drift.

We have not generated an example for this one yet. Everything shown on this site is ours and hosted by us, so where there is no example there is simply nothing to show.

Takes
video, audio
Makes
video, video-edit
Licence
free to use
Download
4 GB
Needs
16 GB RAM

What it costs

Free — downloads once, then runs on your own Mac with no per-image cost.

Where it runs

Downloads once and runs on your own Mac — no key, no account, no per-image cost.

Licence: MIT (code + weights) · commercial OK

← every media model