Run Wan 2.2 in ComfyUI with Just 8GB VRAM – Full Image-to-Video AI Setup + Fixes

If you’re experimenting with AI video workflows and don’t have a high-end GPU — this guide is for you. I recently tested the newly released Wan 2.2 model inside ComfyUI, and got it working smoothly on a machine with only 8GB of VRAM.

🔧 In this post, I cover:

Where to download Wan 2.2 GGUF, Lightx2v LoRA, UMT5 encoder, and VAE

How to organize ComfyUI folders for correct loading

How I fixed a missing sageattention Python module using Gemini CLI

A full image-to-video test with cinematic motion from a still wolf image

Performance tips for low-VRAM systems

🎥 You can watch the full video walkthrough here:

This workflow proves that powerful generative video tools are now accessible to more creators. Let me know if you run into any setup issues — happy to help!

#AI #ComfyUI #Wan22 #VideoGeneration #OpenSource #LowVRAM #GeminiCLI #Tutorial #Steemit #AIworkflow