---
language: en
license: apache-2.0
tags:
  - text-to-video
  - image-to-video
  - wan-2.2
  - character-consistency
  - bf16
  - diffusion-model
  - comfyui
---

# Wan 2.2 Animate 14B BF16 — character-ID-preserving weights mirror

**Organization:** [WindstormLabs](https://huggingface.co/WindstormLabs)  
**Used by:** [SceneMachine](https://huggingface.co/SceneMachine)

SceneMachine is a sub-project of Windstorm Labs. This repo hosts shared AI/ML infrastructure used by SceneMachine and reusable by future Windstorm Labs sub-projects.

## What this is

Mirror of the Wan 2.2 Animate 14B BF16 weights. SceneMachine's character-consistency mode: reference image of a character + text prompt produces video preserving the character's identity. Validated 1.7 min/shot on a 32 GB-VRAM card with the Lightx2v 4-step LoRA from `wan22-loras` (8.3× speedup over baseline). REQUIRES `clip_vision_h.safetensors` from `wan22-encoders` (SigLIP won't work — LayerNorm shape mismatch).

## Upstream source

Primary distribution: https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged

This repo is a **mirror**. License terms of the upstream apply unchanged.

**Primary license owner(s):** Alibaba (Wan 2.2 release)

## Files

| Filename | Size |
|---|---|
| `wan2.2_animate_14B_bf16.safetensors` | 34.55 GB |

**Total: 34.55 GB**

## Related repos

- The full SceneMachine model stack: search the [SceneMachine HF collection](https://huggingface.co/SceneMachine).
- Shared encoders and Wan VAE: [`WindstormLabs/wan22-encoders`](https://huggingface.co/WindstormLabs/wan22-encoders).
- Speed LoRAs: [`WindstormLabs/wan22-loras`](https://huggingface.co/WindstormLabs/wan22-loras).

## Provenance

Mirror created 2026-05-13 from local working copy on the SceneMachine development rig. Hashes preserved via HF's content-addressed storage.

🤖 Repo and README generated by [Claude Code](https://claude.com/claude-code) during a SceneMachine CTO session.