| File | wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors |
| Model | Wan 2.2 T2V A14B (video model, 14B) |
| Precision | FP8 |
| Size | 14.3 GB |
| VRAM to fit | about 18.6 GB |
| Folder | ComfyUI/models/diffusion_models/ |
| Loader node | Load Diffusion Model (UNETLoader) |
| Source | Comfy-Org/Wan_2.2_ComfyUI_Repackaged on Hugging Face |
Download wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors (14.3 GB)
Where to put it
Put the file in ComfyUI/models/diffusion_models/ (portable version: ComfyUI_windows_portable/ComfyUI/models/diffusion_models/). Older guides say models/unet: that folder still works. Press R in ComfyUI or restart it, then pick the file in Load Diffusion Model (UNETLoader). Not for Mac: FP8 and INT8 files do not load on Apple Silicon.
If a workflow says “Value not in list” for that node, the workflow expects a file with another name. Click the node and choose this one.
Which graphics cards it fits
| VRAM | This file | Best file for that card (RTX 40/50) | |
|---|---|---|---|
| 6 GB | Not practical | Short by 12.6 GB | Wan2.2-T2V-A14B-HighNoise-Q4_K_M.gguf (9.7 GB) |
| 8 GB | Offload only | Short by 10.6 GB: runs with part of it in system RAM, slower | Wan2.2-T2V-A14B-HighNoise-Q2_K.gguf (5.3 GB) |
| 10 GB | Offload only | Short by 8.6 GB: runs with part of it in system RAM, slower | Wan2.2-T2V-A14B-HighNoise-Q2_K.gguf (5.3 GB) |
| 12 GB | Offload only | Short by 6.6 GB: runs with part of it in system RAM, slower | Wan2.2-T2V-A14B-HighNoise-Q3_K_M.gguf (7.2 GB) |
| 16 GB | Offload only | Short by 2.6 GB: runs with part of it in system RAM, slower | Wan2.2-T2V-A14B-HighNoise-Q5_K_M.gguf (10.8 GB) |
| 20 GB | Runs well | Fits in VRAM, 1.4 GB to spare | this file |
| 24 GB | Runs well | Fits in VRAM, 5.4 GB to spare | this file |
| 32 GB | Runs well | Fits in VRAM, 13.4 GB to spare | this file |
| 48 GB | Runs well | Fits in VRAM, 29.4 GB to spare | wan2.2_t2v_high_noise_14B_fp16.safetensors (28.6 GB) |
Need = file × 1 + 3.5 GB for the picture and 0.8 GB for Windows and the browser — the same rule as the rest of the site. FP8 files do not load on a Mac (Apple Silicon); use a 16-bit or GGUF file there.
What else you need
This file alone does not make a picture. With it you load:
| Role | File | Size | Folder |
|---|---|---|---|
| Model | wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors high-noise model (early steps) | 14.3 GB | models/diffusion_models |
| Model | wan2.2_t2v_low_noise_14B_fp8_scaled.safetensors low-noise model (late steps) | 14.3 GB | models/diffusion_models |
| Text encoder | umt5_xxl_fp8_e4m3fn_scaled.safetensors | 6.7 GB | models/text_encoders |
| VAE | wan_2.1_vae.safetensors | 0.3 GB | models/vae |
35.6 GB in total. Smaller text-encoder versions and other options are on the Wan 2.2 T2V page.
Other files of Wan 2.2 T2V
| File | Precision | Size | VRAM to fit |
|---|---|---|---|
| wan2.2_t2v_high_noise_14B_fp16.safetensors | 16-bit | 28.6 GB | 32.9 GB |
| Wan2.2-T2V-A14B-HighNoise-Q8_0.gguf | Q8_0 | 15.4 GB | 19.7 GB |
| wan2.2_t2v_high_noise_14B_fp8_scaled.safetensors | FP8 | 14.3 GB | 18.6 GB |
| Wan2.2-T2V-A14B-HighNoise-Q6_K.gguf | Q6_K | 12.0 GB | 16.3 GB |
| Wan2.2-T2V-A14B-HighNoise-Q5_K_M.gguf | Q5_K_M | 10.8 GB | 15.1 GB |
| Wan2.2-T2V-A14B-HighNoise-Q4_K_M.gguf | Q4_K_M | 9.7 GB | 14.0 GB |
| Wan2.2-T2V-A14B-HighNoise-Q3_K_M.gguf | Q3_K_M | 7.2 GB | 11.5 GB |
| Wan2.2-T2V-A14B-HighNoise-Q2_K.gguf | Q2_K | 5.3 GB | 9.6 GB |
Same model, different compression. 8-bit files (FP8, INT8, Q8_0) look almost the same as 16-bit; below Q5 the loss starts to show.
Free workflow with this file
- Wan 2.2 T2V 14B — FP8 — 19 min for an 81-frame 832×480 clip
Drag the .json into ComfyUI. All free workflows →
Want the answer for your exact card? Check your GPU. Every file of every model: ComfyUI model files.
Size and source read from Hugging Face (2026-10-01).