EchoMimic

Free plan LinuxSelf-hostedWeb

What EchoMimic is

EchoMimic is a portrait-video generation project that can animate a portrait using audio, facial landmarks, or both together. Its audio demonstrations cover English, Chinese, and singing. It also supports animation driven by pose alone or by audio and pose, and includes a demonstration for transferring motion between a reference image and a driven video. The project says it was trained with audio and facial landmarks to enable these different driving approaches.

You can run its Python inference scripts or use its Gradio interface. Hosted demonstrations are linked on Hugging Face and ModelScope, and the repository links a community-contributed ComfyUI implementation. Running inference requires pretrained weights from the BadToBest EchoMimic Hugging Face repository. Listed tested environments include CentOS 7.2 or Ubuntu 22.04, CUDA 11.7 or later, Python 3.8, 3.10, or 3.11, and A100 80G, RTX4090D 24G, or V100 16G GPUs.

The project is attributed to the Terminal Technology Department, Alipay, Ant Group, and its paper was accepted by AAAI 2025. Its license is Apache License, Version 2.0. EchoMimic is intended for academic research; the project places responsibility for generated content and actions on users.

Who it suits

EchoMimic suits researchers and developers who want to generate portrait animations with audio, facial landmarks, or pose controls. It is a fit for people able to work with the listed operating systems, CUDA, Python, and tested GPU options, or who want to explore the linked hosted demos.

What it does well

Supports audio, facial landmark, and combined driving modes.
Includes pose-driven animation inference instructions.
Provides Python inference scripts and a Gradio interface.
Links hosted demos on Hugging Face and ModelScope.
Uses the Apache License, Version 2.0.

What to know first

Inference requires downloading pretrained weights separately.
Listed tested setups require CUDA and specified GPU hardware.
The project is intended for academic research.
No support service or response commitment is stated.

Limits

EchoMimic is free. Inference does require pretrained weights and a tested environment with CUDA and a listed GPU.

Getting started

Open the EchoMimic GitHub repository.
Review the installation instructions for a listed operating system, CUDA version, Python version, and GPU.
Download pretrained weights from the BadToBest EchoMimic Hugging Face repository.
Follow the repository instructions to run Python inference scripts or launch the Gradio UI.
Alternatively, open a linked Hugging Face or ModelScope demo.

Verdict

Choose EchoMimic if you need portrait-video generation with audio, landmark, or pose inputs and can use its listed technical environment. Its license and inference code support research use, while users seeking a stated support commitment or a non-research product should look elsewhere.

Questions

How much does EchoMimic cost?

It is free, with a free plan.

What can EchoMimic generate?

It generates portrait videos driven by audio, facial landmarks, or both. The repository also provides pose-driven and audio-and-pose-driven animation instructions.

Which platforms does it support?

The listed platforms are Linux, self-hosted, and web. The project provides Python inference scripts, a Gradio UI, and links to hosted demos.

What license does EchoMimic use?

The repository's LICENSE file contains the Apache License, Version 2.0.

Who makes EchoMimic?

The project page attributes the work to the Terminal Technology Department, Alipay, Ant Group.

Where can users report issues?

The repository provides GitHub Issues as its visible issue-reporting channel.

Facts from EchoMimic's own pages

Audio input: The project page shows audio-driven demos for English, Chinese, and singing. · github.com, 4 Oct 2026
Deployment: The repository provides Python inference scripts and instructions for running a Gradio UI. · github.com, 4 Oct 2026
Hosted demos: The repository links EchoMimic demos on Hugging Face and ModelScope. · github.com, 4 Oct 2026
Installation requirements: The README lists tested CentOS 7.2 or Ubuntu 22.04 environments, CUDA 11.7 or later, Python 3.8, 3.10, or 3.11, and A100, RTX4090D, or V100 GPUs. · github.com, 4 Oct 2026
Integrations: The repository links a ComfyUI implementation contributed by a community member. · github.com, 4 Oct 2026
Intended use: The project states that it is intended for academic research and says users are solely liable for their generated content and actions. · github.com, 4 Oct 2026
Interfaces: The project provides a Gradio UI and links to demos on Hugging Face and ModelScope. · github.com, 4 Oct 2026
Landmark conditioning: The project says it was trained with both audio and facial landmarks to support those different driving modes. · antgroup.github.io, 4 Oct 2026
Landmark control: EchoMimic supports landmark-driven animation and audio combined with selected landmarks. · github.com, 4 Oct 2026
License: The repository's LICENSE file contains the Apache License, Version 2.0. · github.com, 4 Oct 2026
Maker: The project page attributes the work to the Terminal Technology Department, Alipay, Ant Group. · antgroup.github.io, 4 Oct 2026
Motion alignment: The repository includes a demo for aligning motion between a reference image and a driven video. · github.com, 4 Oct 2026
Pose control: The repository includes inference instructions for audio-and-pose-driven and pose-driven animation. · github.com, 4 Oct 2026
Publication: The repository says the EchoMimic paper was accepted by AAAI 2025. · github.com, 4 Oct 2026
Purpose: EchoMimic generates portrait videos from audio, facial landmarks, or a combination of audio and selected facial landmarks. · antgroup.github.io, 4 Oct 2026
Requirements: The repository lists tested environments as CentOS 7.2 or Ubuntu 22.04 with CUDA 11.7 or later, Python 3.8, 3.10, or 3.11, and tested GPUs A100 80G, RTX4090D 24G, or V100 16G. · github.com, 4 Oct 2026
Research results: The project page says EchoMimic was compared with alternative algorithms on public and collected datasets and showed superior quantitative and qualitative performance. · antgroup.github.io, 4 Oct 2026
Support: The repository provides GitHub Issues as its visible issue-reporting channel. · github.com, 4 Oct 2026
Support and service: The project pages provide code, installation instructions, and demo links; they do not state a support service or response commitment. · github.com, 4 Oct 2026
Weights: Inference setup requires downloading pretrained weights from the BadToBest EchoMimic Hugging Face repository. · github.com, 4 Oct 2026

Where EchoMimic stands on Freedom251

AI Video Lip Sync Tools#6 of 28