https://github.com/antgroup/echomimic_v3
[AAAI 2026] EchoMimicV3: 1.3B Parameters are All You Need for Unified Multi-Modal and Multi-Task Human Animation
https://github.com/antgroup/echomimic_v3
audio-driven-body-animation audio-driven-portrait-animations human-animation video-generation
Last synced: 6 months ago
JSON representation
[AAAI 2026] EchoMimicV3: 1.3B Parameters are All You Need for Unified Multi-Modal and Multi-Task Human Animation
- Host: GitHub
- URL: https://github.com/antgroup/echomimic_v3
- Owner: antgroup
- License: apache-2.0
- Created: 2025-07-24T02:20:55.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2026-01-30T05:22:16.000Z (6 months ago)
- Last Synced: 2026-01-30T21:42:23.220Z (6 months ago)
- Topics: audio-driven-body-animation, audio-driven-portrait-animations, human-animation, video-generation
- Language: Python
- Homepage: https://antgroup.github.io/ai/echomimic_v3/
- Size: 50.8 MB
- Stars: 739
- Watchers: 8
- Forks: 83
- Open Issues: 21
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
- awesome-ai - EchoMimicV3
README
[简体中文](https://github.com/antgroup/echomimic_v3/blob/main/README_zh.md) | English
EchoMimicV3: 1.3B Parameters are All You Need for Unified Multi-Modal and Multi-Task Human Animation
Terminal Technology Department, Alipay, Ant Group.
1Core Contributor
2Corresponding Authors
## 🚀 EchoMimic Series
* EchoMimicV1: Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning. [GitHub](https://github.com/antgroup/echomimic)
* EchoMimicV2: Towards Striking, Simplified, and Semi-Body Human Animation. [GitHub](https://github.com/antgroup/echomimic_v2)
* EchoMimicV3: 1.3B Parameters are All You Need for Unified Multi-Modal and Multi-Task Human Animation. [GitHub](https://github.com/antgroup/echomimic_v3)
## 📣 Updates
* [2026.01.22] 🔥 We update our EchoMimicV3-Flash on [Huggingface](https://huggingface.co/BadToBest/EchoMimicV3/tree/main/echomimicv3-flash-pro).
- 🚀 8-step High-quality Generation.
- 🧩 No Face Mask required.
- 💾 12G VRAM Requirement.
- ✅ Supports up to 768×768 Resolution.
* [2025.11.09] 🔥 EchoMimicV3 is accepted by AAAI 2026.
* [2025.08.21] 🔥 EchoMimicV3 gradio demo on [modelscope](https://modelscope.cn/studios/BadToBest/EchoMimicV3) is ready.
* [2025.08.12] 🔥🚀 **12G VRAM is All YOU NEED to Generate Video**. Please use this [GradioUI](https://github.com/antgroup/echomimic_v3/blob/main/app_mm.py). Check the [tutorial](https://www.bilibili.com/video/BV1W8tdzEEVN) from @[gluttony-10](https://github.com/gluttony-10). Thanks for the contribution.
* [2025.08.12] 🔥 EchoMimicV3 can run on **16G VRAM** using [ComfyUI](https://github.com/smthemex/ComfyUI_EchoMimic). Thanks @[smthemex](https://github.com/smthemex) for the contribution.
* [2025.08.09] 🔥 We release our [models](https://modelscope.cn/models/BadToBest/EchoMimicV3) on ModelScope.
* [2025.08.08] 🔥 We release our [codes](https://github.com/antgroup/echomimic_v3) on GitHub and [models](https://huggingface.co/BadToBest/EchoMimicV3) on Huggingface.
* [2025.07.08] 🔥 Our [paper](https://arxiv.org/abs/2507.03905) is in public on arxiv.
## 🌅 Gallery
### Chinese Driven Audio
For more demo videos, please refer to the [project page](https://antgroup.github.io/ai/echomimic_v3/)
## Quick Start
### Environment Setup
- Tested System Environment: Centos 7.2/Ubuntu 22.04, Cuda >= 12.1
- Tested GPUs: A100(80G) / RTX4090D (24G) / V100(16G)
- Tested Python Version: 3.10 / 3.11
### 🛠️Installation for Windows
##### Please use the [one-click installation package](https://pan.baidu.com/share/init?surl=cV7i2V0wF4exDtKjJrAUeA) (passport: glut) to get started quickly for Quantified version.
### 🛠️Installation for Linux
#### 1. Create a conda environment
```
conda create -n echomimic_v3 python=3.10
conda activate echomimic_v3
```
#### 2. Other dependencies
```
pip install -r requirements.txt
```
### 🧱Model Preparation
| Models | Download Link | Notes |
| --------------|-------------------------------------------------------------------------------|-------------------------------|
| Wan2.1-Fun-V1.1-1.3B-InP | 🤗 [Huggingface](https://huggingface.co/alibaba-pai/Wan2.1-Fun-V1.1-1.3B-InP) | Base model
| wav2vec2-base | 🤗 [Huggingface](https://huggingface.co/facebook/wav2vec2-base-960h) | Audio encoder for preview
| chinese-wav2vec2-base | 🤗 [Huggingface](https://modelscope.cn/models/TencentGameMate/chinese-wav2vec2-base) | Audio encoder for Flash
| EchoMimicV3-preview | 🤗 [Huggingface](https://huggingface.co/BadToBest/EchoMimicV3) | preview weights
| EchoMimicV3-preview | 🤗 [ModelScope](https://modelscope.cn/models/BadToBest/EchoMimicV3) | preview weights
| EchoMimicV3-Flash | 🤗 [Huggingface](https://huggingface.co/BadToBest/EchoMimicV3/tree/main/echomimicv3-flash-pro) | Flash weights
-- The **weights** of EchoMimicV3-flash-pro is organized as follows.
```
./flash-pro/
├── Wan2.1-Fun-V1.1-1.3B-InP
├── chinese-wav2vec2-base
└── transformer
└── diffusion_pytorch_model.safetensors
```
-- The **weights** is of EchoMimicV3-preview organized as follows.
```
./preview/
├── Wan2.1-Fun-V1.1-1.3B-InP
├── wav2vec2-base-960h
└── transformer
└── diffusion_pytorch_model.safetensors
```
### 🔑 Quick Inference for EchoMimicV3-flash-pro
```
bash run_flash_pro.sh
```
### 🔑 Quick Inference for EchoMimicV3-preview
```
python infer_preview.py
```
For Quantified GradioUI version for EchoMimicV3-preview:
```
python app_mm.py
```
**images, audios, masks and prompts are provided in `datasets/echomimicv3_demos`**
#### Tips
- Audio CFG: Audio CFG `audio_guidance_scale` works optimally between 1.8~2. Increase the audio CFG value for better lip synchronization, while decreasing the audio CFG value can improve the visual quality.
- Text CFG: Text CFG `guidance_scale` works optimally between 3~6. Increase the text CFG value for better prompt following, while decreasing the text CFG value can improve the visual quality.
- TeaCache: The optimal range for `teacache_threshold` is between 0~0.1.
- Sampling steps: 5 steps for talking head, 15~25 steps for talking body.
- Long video generation: If you want to generate a video longer than 138 frames, you can use Long Video CFG.
- Try setting `partial_video_length` to 81, 65 or smaller to reduce VRAM usage.
## 📒 Citation
If you find our work useful for your research, please consider citing the paper :
```
@misc{meng2025echomimicv3,
title={EchoMimicV3: 1.3B Parameters are All You Need for Unified Multi-Modal and Multi-Task Human Animation},
author={Rang Meng, Yan Wang, Weipeng Wu, Ruobing Zheng, Yuming Li, Chenguang Ma},
year={2025},
eprint={2507.03905},
archivePrefix={arXiv}
}
```
## Reference
- Wan2.1: https://github.com/Wan-Video/Wan2.1/
- VideoX-Fun: https://github.com/aigc-apps/VideoX-Fun/
## 📜 License
The models in this repository are licensed under the Apache 2.0 License. We claim no rights over the your generated contents,
granting you the freedom to use them while ensuring that your usage complies with the provisions of this license.
You are fully accountable for your use of the models, which must not involve sharing any content that violates applicable laws,
causes harm to individuals or groups, disseminates personal information intended for harm, spreads misinformation, or targets vulnerable populations.
## 🌟 Star History
[](https://www.star-history.com/#antgroup/echomimic_v3&Date)
