## GingerLabs Wan 2.2 T2V A14BA GingerLabs-tuned derivative of [Wan-AI/Wan2.2-T2V-A14B](<a target="_blank" rel="ugc" href="https://huggingface.co/Wan-AI/Wan2.2-T2V-A14B">https://huggingface.co/Wan-AI/Wan2.2-T2V-A14B</a>), packaged as matching HighNoise and LowNoise diffusion-model pairs for ComfyUI.Available formats:- GGUF <code>Q4_K_M</code>- GGUF <code>Q5_K_M</code> — recommended balance- GGUF <code>Q6_K</code>- GGUF <code>Q8_0</code>- Scaled <code>FP8 E4M3FN</code>No acceleration LoRA is required.### Important: download both filesEach version requires its matching HighNoise and LowNoise files. Keep both files at the same quantization level and do not swap their roles.Example:- <code>GingerLabs_Wan2.2_T2V_A14B_High_Q5_K_M.gguf</code>- <code>GingerLabs_Wan2.2_T2V_A14B_Low_Q5_K_M.gguf</code>Place both files in:<code>ComfyUI/models/diffusion_models/</code>### Requirements- [ComfyUI-WanVideoWrapper](<a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-WanVideoWrapper">https://github.com/kijai/ComfyUI-WanVideoWrapper</a>) 1.3.6 or newer- ComfyUI-VideoHelperSuite- <code>Wan2_1_VAE_bf16.safetensors</code>- Exactly <code>nsfw_wan_umt5-xxl_bf16.safetensors</code> from [NSFW-API/NSFW-Wan-UMT5-XXL](<a target="_blank" rel="ugc" href="https://huggingface.co/NSFW-API/NSFW-Wan-UMT5-XXL/tree/main?not-for-all-audiences=true">https://huggingface.co/NSFW-API/NSFW-Wan-UMT5-XXL/tree/main?not-for-all-audiences=true</a>)The scaled-FP8 version of that UMT5 encoder is not supported by the included workflow.Place the supporting files in:- Text encoder: <code>ComfyUI/models/text_encoders/</code>- VAE: <code>ComfyUI/models/vae/</code>### Recommended starting settings- Total steps: <code>8</code>- HighNoise range: <code>0–4</code>- LowNoise range: <code>4–end</code>- Sampler: <code>Euler</code>- Shift: <code>5.0</code>- CFG: <code>1.0</code>- Resolution: <code>720 × 1280</code>- Frames: <code>33</code>- FPS: <code>16</code>This means 8 total denoising steps, split as four HighNoise and four LowNoise steps—not 8+8.At CFG 1.0, negative conditioning is skipped, so the negative prompt does not affect sampling.For the WanVideoWrapper model loaders, leave <code>quantization</code> set to <code>disabled</code>. The wrapper loads GGUF directly and detects the scaled-FP8 format. <code>SDPA</code> is the portable attention default; SageAttention is optional.### Scope- Text-to-video only- Not intended for I2V, TI2V, FLF2V, S2V, Animate, or VACE- No trigger words- Both HighNoise and LowNoise files are required- Frame counts should follow <code>4n+1</code>The files passed structural and load/unload validation. A comparative visual-quality or performance benchmark between quantization levels has not been performed.### Downloads and workflow- [Reference ComfyUI workflow](<a target="_blank" rel="ugc" href="https://huggingface.co/GingerLabsPlatform/GingerLabs-Wan2.2-T2V-A14B/resolve/main/workflows/GingerLabs_T2V_Workflow.json">https://huggingface.co/GingerLabsPlatform/GingerLabs-Wan2.2-T2V-A14B/resolve/main/workflows/GingerLabs_T2V_Workflow.json</a>)- [Full model card, files, and checksums](<a target="_blank" rel="ugc" href="https://huggingface.co/GingerLabsPlatform/GingerLabs-Wan2.2-T2V-A14B">https://huggingface.co/GingerLabsPlatform/GingerLabs-Wan2.2-T2V-A14B</a>)

Files with SHA256: 0bd43225d437002b256a3bf166796917aa24d5aea2b1b45624db8abf63d54967