Skip to main content

InfiniteTalk online

InfiniteTalk is MeiGen-AI's open model that turns a still photo and a voice track into a lip-synced talking video. Run it below without installing anything: add a portrait and up to 15 seconds of voice, and pay per second.

  • By MeiGen-AI
  • Updated 2026-11-20

What is InfiniteTalk?

InfiniteTalk is an audio-driven video model from MeiGen-AI. Its GitHub repository describes it as audio-driven video generation for "sparse-frame video dubbing", and its news list dates the release of the technical report, weights and code to August 19, 2025, the same day the paper was submitted to arXiv. The project page lists Meituan among the authors' affiliations, alongside universities and research institutes.

The README says the model syncs not only the lips but also the head, body and expressions to the audio, and that it builds on the Wan2.1 image-to-video model, with an audio encoder that reads the voice.

From a photo or from a video

InfiniteTalk works two ways. Its project page shows it dubbing an existing video, keeping the identity, background and camera movement while the person speaks the new audio, and its README says it can also animate a single image. The README lists 480p and 720p, and a separate checkpoint for several people. It also notes that from a single image, results stay good for up to about a minute, after which colors drift more.

On MakeMidas, InfiniteTalk animates one portrait (JPG, PNG or WebP) to a voice track of up to 15 seconds, at 480p or 720p. You can add a short direction for expression and movement. For longer videos, Kling AI Avatar 2.0 takes up to 5 minutes of voice; both are on the AI talking avatar page.

A script instead of a recording

InfiniteTalk speaks the audio you give it, so the voice is yours to choose. Record yourself, or turn a script into a voiceover with text to speech, download it, and upload it with the photo. For an AI spokesperson or AI actor who does not exist, make the portrait first in the AI image generator.

InfiniteTalk on YouTube

Demos and reviews by other people, for a second opinion.

InfiniteTalk in ComfyUI Tutorial – The Next Level of AI Talking Avatar!Benji's AI Playground · 2025-08-22
InfiniteTalk Deep Dive: A Step-by-Step ComfyUI Workflow for AI Talking AvatarsCodebreakers · 2025-08-28
ComfyUI Tutorial Series Ep 60 Infinite Talk (Audio-Driven Talking AI Characters)pixaroma · 2025-09-02

Questions

Is InfiniteTalk free?

The model is open: MeiGen-AI publishes the code on GitHub and the weights on Hugging Face, to run on your own hardware. On MakeMidas it costs credits per second of the voice track, as the table above shows; new accounts get 50 credits after verifying their email. See pricing for plans and packs.

How long can an InfiniteTalk video be?

The README describes unlimited-length generation, with the best results from a single image up to about a minute. On MakeMidas, InfiniteTalk takes up to 15 seconds of voice; for up to 5 minutes, use Kling AI Avatar 2.0.

Can InfiniteTalk lip-sync an existing video?

The model can: it was built for dubbing video as well as animating photos. On MakeMidas it animates a photo only; video-to-video lip sync is not offered.

Can I use InfiniteTalk videos commercially?

Check before you do. The GitHub repository and Hugging Face model card license the models under Apache 2.0, but the project page says generated content is for academic use only. Our licenses page explains how commercial use follows each model's own terms.

Is there a InfiniteTalk API?

Yes. MakeMidas offers InfiniteTalk through a REST API, an MCP server and a CLI, paid with credits at the prices in the table above. Create an API key in Settings → API keys; the developer guide has every endpoint.

Sources

From the maker

  1. InfiniteTalk on GitHub · MeiGen-AI, 2025-08-19
  2. InfiniteTalk: Audio-driven Video Generation for Sparse-Frame Video Dubbing · arXiv, Shaoshu Yang et al., 2025-08-19
  3. InfiniteTalk model card · MeiGen-AI on Hugging Face
  4. InfiniteTalk project page · MeiGen-AI

Try InfiniteTalk now

Write a prompt or add a picture, and pay only for what you make.