AI model · Video

MuseTalk

MuseTalk, published by Tencent (Lyra Lab), is a real-time, audio-driven lip-sync model.

  • Portrait
  • Expressif
  • Talking head
  • Video
Provider Tencent

MuseTalk

VideoHavincy
Provider
Tencent
Type
Video
Cost
From 23 credits

Overview

What is MuseTalk?

MuseTalk, published by Tencent (Lyra Lab), is a real-time, audio-driven lip-sync model. It animates only the mouth area to match your voice, preserving the rest of the source video.

In Havincy, MuseTalk syncs an existing video to an audio track at very low cost. It’s a good choice for communication avatars and simple localization.

Key strengths

  • Portrait
  • Expressif
  • Talking head

Specs

MuseTalk: specifications

Publisher
Tencent
Generation modes
lip sync (video + voice)
Available since
April 11, 2024
License
Commercial use allowed
Cost in Havincy
From 23 credits

Specs taken from the publisher’s public model page. See the documentation on fal.ai

Use cases

What you can create with MuseTalk

Communication avatars

A recurring face announcing your news.

Simple localization

Translated versions of short videos.

Prototypes

Test a spoken script before a more polished production.

Internal content

Team announcements and messages with a virtual presenter.

Prompt examples

Prompts to try with MuseTalk

Paste an example into Havincy, then adapt the subject, style and format to your project.

  • Sync the mouth of this video to the announcement audio.

  • Have our avatar say this new message.

  • Dub this welcome message into Spanish.

Use cases

Compatible skills

Use this model with the following Havincy skills.

How it works

How to create with this model

  1. Describe the desired result in a Havincy conversation.
  2. Add a reference when the style, product, or character needs to be preserved.
  3. Generate, then ask for a variation or improvement without starting over.

FAQ

Frequently asked questions about MuseTalk

How much does MuseTalk cost?

fal bills it by compute time, with no published price. Havincy applies an estimate of about 23 credits for 8 seconds.

Is MuseTalk open source?

Yes, MuseTalk is open source (GitHub). In Havincy, no install is needed.

MuseTalk or Wav2Lip?

MuseTalk usually produces a sharper, more natural mouth than older models like Wav2Lip, while staying fast.

Ready to create with MuseTalk?

Describe your idea in Havincy, then refine the result in the conversation.

Create in Havincy