alt.freeweights.video / Wan2.2-Animate-2-14B
From: Wan-AI <huggingface.co/Wan-AI>Newsgroups: alt.freeweights.videoSubject:

Wan2.2-Animate-2-14B

Date: 2026-11-15Commit: 6e8f197Signed: 2DC28CC7B26A283B ✓

Model. Safetensors weights, 82.5 GB in all. Published by Wan-AI under Apache-2.0.

video82.5GBWan2.2-Animate-2-14Bsafetensorspermissive
FactsParamsnot statedSize82.5 GBFormatssafetensorsTasknot taggedLicenseApache-2.0From sourcepermissive
$ fw get Wan-AI/Wan2.2-Animate-2-14B
# manifest sha256 8666917bae8c… · 82.5 GB · 16 files
$ fw verify ./Wan2.2-Animate-2-14B Wan-AI/Wan2.2-Animate-2-14B
From sourceHosted (signed)Wan-AI/Wan2.2-Animate-2-14BLast check: 2026-11-15 04:31 UTC

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Works with

Runs with

Ranked by fit to this model (safetensors), then by GitHub stars.

Overview

Size
82.5 GB
Weights
65.6 GB (safetensors)
Parameters
not stated
Formats
safetensors
Task
not tagged
License
Apache-2.0 (permissive)
Files
16, each with a sha256
Publisher
Wan-AI
Added
2026-11-15
Section
alt.freeweights.video
Lineage

Original release from the publisher.

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 04:31 UTCpassed

Model card

From the publisher at commit 6e8f197. Shown as plain text with links; images open at the publisher.

Wan-Animate-2: Pushing the Application#

Boundaries of Character Animation Models

Guangyuan Wang^∗^, Li Hu^†^, Dechao Meng^∗^, Zhongyi Zhang^∗^, Peng Zhang^∗^, Mingyang Huang, Ruoshi Zhang, Ke Sun, Zhe Zhang, Xingjun Wang, Gang Cheng, Hai Xu, Bang Zhang*^‡^

∗Core Contribution &nbsp; †Project leaders &nbsp; ‡Sponsor

Version &nbsp; License &nbsp; arXiv &nbsp; Project Page &nbsp; Demo

📝 Introduction#

We present Wan-Animate-2, a novel end-to-end character animation framework that directly consumes driving videos in a redesigned Diffusion Transformer, which achieves high-fidelity motion generation and strong identity preservation by eliminating intermediate motion extractors. We further add text-driven viewpoint control to decouple the output camera perspective from the driving video. In addition, we develop Wan-Animate-2-Lite, an efficient variant that reduces inference latency to real-time thresholds for streaming character animation.

Image: architecture

Release Notes:

  • August 07, 2026: 🎉 We release the Wan-Animate-2 inference scripts.
  • August 07, 2026: 🎉 We release the Wan-Animate-2 Base model weights.
  • August 07, 2026: 🎉 We release the Wan-Animate-2 Distillation model weights.

📑 Todo List#

  • Wan-Animate-2 Character Animation
    • [x] Inference code of Wan-Animate-2
    • [x] Checkpoints of Wan-Animate-2
    • [x] Diffusers integration
    • [x] DiffSynth-Studio integration
    • [x] ComfyUI integration

🚀 Quick Start#

Installation#
  1. Clone this repo:
git clone --recursive https://github.com/Wan-Video/Wan-Animate-2.git
  1. Create python environment and install torch:
conda create -n wan_animate_2 python==3.11 -y
pip install torch==2.7.0 torchvision==0.22.0 torchaudio==2.7.0 --index-url https://download.pytorch.org/whl/cu126
  1. Install dependencies (please refer to requirements.txt):
pip install -r requirements.txt
  1. Install flash_attn:
pip install flash-attn --no-build-isolation
  1. Install this repository as a local editable package:
pip install -e .
Model Download#
ModelDownload Links
Wan-Animate-2🤗 HuggingFace 🤖 ModelScope

Download models using huggingface-cli:

pip install "huggingface_hub[cli]"
huggingface-cli download Wan-AI/Wan2.2-Animate-2-14B --local-dir ./ckpts/

Download models using modelscope-cli:

pip install modelscope
modelscope download --model Wan-AI/Wan2.2-Animate-2-14B --local_dir ./ckpts/
Inference#

Before inference, use an LLM model (e.g. Qwen3.7-Plus) with the prompt below to get the image caption, then use it as the prompt for Wan-Animate-2.

用中文客观描述图片中的内容,包括以下要点:人物外观描述,不描述动作行为。 背景描述,忽略主观评价和情绪推测。 下面给出描述范例,必须遵循这个范式,不要输出额外的符号: 人物外观描述:穿着一件浅蓝色的校服衬衫,领口和袖口有白色边饰。胸前有一个圆形徽章。 背景描述:背景为明亮、整洁的教室或办公室,氛围安静有序。

Note: The default settings in this repository are tuned for 8× A800 GPUs and support 720P video generation. We have also tested 480P generation on 2× A800 GPUs. If your hardware setup is different, please adjust the parallel configs in the YAML files accordingly.

  • Wan-Animate-2 Base:
  export PYTHONPATH="$(pwd)"
  cd infer
  python wan_animate_2_demo.py \
    --prompt "人物外观描述:一只银灰色虎斑纹的小猫,拥有圆润的脸庞、竖立的耳朵和巨大的圆形眼睛。它身穿一套深蓝色的制服套装,包括一件带有金色纽扣的西装外套和一条百褶裙。外套里面搭配着白色衬衫,领口处系着一个红色的蝴蝶结,袖口露出白色的衬衫边缘。背景描述:背景为纯白色,光线均匀明亮,无其他杂物或装饰。" \
    --refer-img-file ../examples/demo1/reference.png  \
    --refer-video-file ../examples/demo1/template.mp4 \
    --config ./wan_animate_2.yaml
  • Wan-Animate-2 Distillation:
  export PYTHONPATH="$(pwd)"
  cd infer
  python wan_animate_2_demo.py \
    --prompt "人物外观描述:一只银灰色虎斑纹的小猫,拥有圆润的脸庞、竖立的耳朵和巨大的圆形眼睛。它身穿一套深蓝色的制服套装,包括一件带有金色纽扣的西装外套和一条百褶裙。外套里面搭配着白色衬衫,领口处系着一个红色的蝴蝶结,袖口露出白色的衬衫边缘。背景描述:背景为纯白色,光线均匀明亮,无其他杂物或装饰。" \
    --refer-img-file ../examples/demo1/reference.png  \
    --refer-video-file ../examples/demo1/template.mp4 \
    --config ./wan_animate_2_distillation.yaml \
    --sample_guide_scale 1.0 
    --step 10

🧨 Diffusers Inference#

Wan-Animate-2 is supported by the 🤗 diffusers library (see PR #14412).

Install diffusers from source (until the next release):

pip install git+https://github.com/huggingface/diffusers.git
pip install flash-attn --no-build-isolation
  • Wan-Animate-2 Base:
  import torch
  from diffusers import WanAnimate2Pipeline
  from diffusers.utils import export_to_video, load_image

  pipe = WanAnimate2Pipeline.from_pretrained(
      "Wan-AI/Wan2.2-Animate-2-14B-Diffusers", torch_dtype=torch.bfloat16
  ).to("cuda")

  output = pipe(
      image=load_image("../examples/demo1/reference.png"),
      driving_video="../examples/demo1/template.mp4",
      prompt="人物外观描述:一只银灰色虎斑纹的小猫,拥有圆润的脸庞、竖立的耳朵和巨大的圆形眼睛。它身穿一套深蓝色的制服套装,包括一件带有金色纽扣的西装外套和一条百褶裙。外套里面搭配着白色衬衫,领口处系着一个红色的蝴蝶结,袖口露出白色的衬衫边缘。背景描述:背景为纯白色,光线均匀明亮,无其他杂物或装饰。",
      height=800,
      width=640,
      num_inference_steps=40,
  )

  export_to_video(output.frames[0], "output.mp4", fps=24)
  • Wan-Animate-2 Distillation (10 steps, no CFG):
  pipe = WanAnimate2Pipeline.from_pretrained(
      "Wan-AI/Wan2.2-Animate-2-14B-Distilled-Diffusers", torch_dtype=torch.bfloat16
  ).to("cuda")

  output = pipe(
      image=load_image("../examples/demo1/reference.png"),
      driving_video="../examples/demo1/template.mp4",
      prompt="人物外观描述:一只银灰色虎斑纹的小猫,拥有圆润的脸庞、竖立的耳朵和巨大的圆形眼睛。它身穿一套深蓝色的制服套装,包括一件带有金色纽扣的西装外套和一条百褶裙。外套里面搭配着白色衬衫,领口处系着一个红色的蝴蝶结,袖口露出白色的衬衫边缘。背景描述:背景为纯白色,光线均匀明亮,无其他杂物或装饰。",
      num_inference_steps=10,
      guidance_scale=1.0,        # no classifier-free guidance
      flow_solver="euler",       # Euler scheduler for distilled model
  )

  export_to_video(output.frames[0], "output.mp4", fps=24)

🤖 Gradio Demo#

Try the online demo: ModelScope Studio

Run locally:

  • Wan-Animate-2 Base:
  cd infer
  python wan_animate_2_gradio.py
  • Wan-Animate-2 Distillation:
  cd infer
  python wan_animate_2_gradio_distillation.py

📜 Citation#

If you find this work helpful, please consider citing:

@article{wang2026wananimate2,
  title   = {Wan-Animate-2: Real-Time End-to-End Character Animation via Diffusion Transformer},
  author  = {Wang, Guangyuan and Hu, Li and Meng, Dechao and Zhang, Zhongyi and Zhang, Peng and
             Huang, Mingyang and Zhang, Ruoshi and Sun, Ke and Zhang, Zhe and
             Wang, Xingjun and Cheng, Gang and Zhang, Bang},
  journal = {arXiv preprint arXiv:TODO.06009},
  year    = {2026},
  url     = {https://arxiv.org/abs/TODO.06009}
}

⚖️ License#

This project is licensed under the Apache License 2.0.

Files

16 files · 82.5 GBWeights 65.6 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 16 files with their sha256
FileSizesha256
.gitattributes2.31 KBc36cc510c4c9c345b7be527776ef8327cd95441a950ae920b5fd24865f637602
README.md9.65 KBc24072d3d7eedcc15cb5d10317c742e1abe80cc49be2218b9cd57716814a418c
assets/pipeline-v2.png602 KB7f858fbaaefb0039ee1442066ffb0985fa8eef24500115437e3c8dd0e21ac0d1
videomodel/Wan-AI/models_clip_open-clip-xlm-roberta-large-vit-huge-14.pth4.77 GB628c9998b613391f193eb67ff68da9667d75f492911e4eb3decf23460a158c38
videomodel/Wan-AI/models_t5_umt5-xxl-enc-bf16.pth11.4 GB7cace0da2b446bbbbc57d031ab6cf163a3d59b366da94e5afe36745b746fd81d
videomodel/Wan-AI/umt5-xxl/special_tokens_map.json6.62 KB7b8a9f5040adb67b5805abdfd42c1f8d0f3d0e711f10726580eb3789cd0ad61d
videomodel/Wan-AI/umt5-xxl/spiece.model4.55 MBe3909a67b780650b35cf529ac782ad2b6b26e6d1f849d3fbb6a872905f452458
videomodel/Wan-AI/umt5-xxl/tokenizer.json16.8 MB6e197b4d3dbd71da14b4eb255f4fa91c9c1f2068b20a2de2472967ca3d22602b
videomodel/Wan-AI/umt5-xxl/tokenizer_config.json61.7 KBed9a3a8b0faa71a70a32847e0435fe036e6e112d4df4edb7bb48a921e344dc05
videomodel/Wan-AI/vae.pth784 MB93154080ec8284f1b6a6bf222a252c9f97c6b887f45d26e4099c0e55fe470cbf
videomodel/Wan-AI/xlm-roberta-large/sentencepiece.bpe.model5.07 MBcfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
videomodel/Wan-AI/xlm-roberta-large/special_tokens_map.json280 B06e405a36dfe4b9604f484f6a1e619af1a7f7d09e34a8555eb0b77b66318067f
videomodel/Wan-AI/xlm-roberta-large/tokenizer.json17.1 MB62c24cdc13d4c9952d63718d6c9fa4c287974249e16b7ade6d5a85e7bbb75626
videomodel/Wan-AI/xlm-roberta-large/tokenizer_config.json418 Befb5c0d09722e5fe59a462cd2a9976ee216d55b037597d997cd3fe833216da15
wan_animate_2/wan_animate_2_bf16.safetensors32.8 GB48abc389b8d9bba17a7f54a1cd7f1286fd3e3e0e292ddf756721aee324aede09
wan_animate_2/wan_animate_2_bf16_distillation.safetensors32.8 GB66161359fc58a1d6c46e14fe2d81c881ccaca440757b1905e01a91902bff29d2
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
6e8f1973bf pinned now2026-11-15 04:31 UTC16 files · 82.5 GB

Use it

$ fw get Wan-AI/Wan2.2-Animate-2-14B
# manifest sha256 8666917bae8c… · 82.5 GB · 16 files
$ fw verify ./Wan2.2-Animate-2-14B Wan-AI/Wan2.2-Animate-2-14B
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/Wan-AI/Wan2.2-Animate-2-14B.json --dir ./Wan2.2-Animate-2-14B $ python3 fw.py verify ./Wan2.2-Animate-2-14B ./Wan-AI__Wan2.2-Animate-2-14B.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download Wan-AI/Wan2.2-Animate-2-14B --revision 6e8f1973bf0abc2aafd517992e8b6d88c3c46e69 --local-dir ./Wan2.2-Animate-2-14B
Packaged for ComfyUI. Most image, video and music models run in ComfyUI . Check the publisher's page for the workflow. Freeweights does not host it.

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
c24072d3d7eedcc15cb5d10317c742e1abe80cc49be2218b9cd57716814a418c
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: Wan-AI <huggingface.co/Wan-AI>Newsgroups: alt.freeweights.videoSubject: Wan2.2-Animate-2-14BDate: 2026-11-15Commit: 6e8f197Signed: 2DC28CC7B26A283B ✓
Upstream
Wan-AI/Wan2.2-Animate-2-14B
Pinned commit
6e8f1973bf0abc2aafd517992e8b6d88c3c46e69
Manifest
Wan2.2-Animate-2-14B.json · .minisig
Manifest sha256
8666917bae8c680be325075a35a33e0864f98897c2a830a85c9cbde4498473aa
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 04:31 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/Wan-AI/Wan2.2-Animate-2-14B.json $ curl -O https://freeweights.allternit.com/manifests/Wan-AI/Wan2.2-Animate-2-14B.json.minisig $ minisign -Vm Wan2.2-Animate-2-14B.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

No related models in this catalog.Nothing here lists this model as its parent, and it names no parent.
Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download Wan-AI/Wan2.2-Animate-2-14B
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. Wan2.2-Animate-2-14B.json. Provenance. Takedown history: none.
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: Bernini-R
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: Causal-Forcing-Framewise
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: HuMo
From: bytedance-research <huggingface.co/bytedance-research>Subject: HuMo
From: genmo <huggingface.co/genmo>Subject: mochi-1-preview
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: Mochi-Preview