Wan 2.1 image-to-video (14B)

Animate a still image with Wan 2.1 14B I2V.

Estimated VRAM

~16.4 GB — ⚠️ ESTIMATE (unverified real-world usage). Estimate, not a measured figure: weight-size floor = wan2.1_i2v_480p_14B_fp8_e4m3fn.safetensors (16.4 GB, sizes verified on Hugging Face/GitHub). Activations add more; ComfyUI offloading can reduce it. Source, checked 2026-10-13.

Models required

• wan2.1_i2v_480p_14B_fp8_e4m3fn.safetensors — 16.397 GB ✅ verified 2026-10-13 (listing) · sha256 0ca75338e7a47ca7cacddb7e626647e65829c497387f718ecb6ea0bae456944a → models/diffusion_models
• clip_vision_h.safetensors — 1.264 GB ✅ verified 2026-10-13 (listing) · sha256 64a7ef761bfccbadbaa3da77366aac4185a6c58fa5de5f589b42a65bcc21f161 → models/clip_vision

Tested GPUs

No verified test reports yet.

Nodes

UNETLoader CLIPVisionLoader CLIPVisionEncode WanImageToVideo KSampler

Common errors & fixes

OOM at 720p

Use the 480p model and resolution. ⚠️ unverified: Fix is general guidance; not confirmed against the cited source.

Low-VRAM tips

⚠️ General guidance, not individually verified.

• Use GGUF quantized weights via ComfyUI-GGUF (city96)
• 480p only

Source & credit

Workflow by comfyanonymous (ComfyUI_examples) / Wan team. Get it from the original: https://comfyanonymous.github.io/ComfyUI_examples/wan/. License: Wan 2.1: Apache-2.0.