My personal basic and simple Wan2.1/Wan2.2 I2V workflows based on ComfyUI native one — Wan2.2+UltimateSD_Upscale

This ComfyUI-native Wan2.1/Wan2.2 image-to-video workflow is for turning an image into video; it lists nodes for image loading, text encoding, video creation, image preview, and video saving.

The listed required node packs are ComfyUI, ComfyUI-KJNodes, and ComfyUI-SkyReels-A2.

The listed image and text nodes are LoadImage, ImageScale, and CLIPTextEncode; video and preview nodes are WanImageToVideo, CreateVideo, SaveVideo, and PreviewImage.

The listed base model is Wan Video 2.2 I2V-A14B.

The listed model filenames are lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank64_bf16.safetensors, umt5_xxl_fp16.safetensors, wan_2.1_vae.safetensors, wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors, and wan2.2_i2v_low_noise_14B_fp16.safetensors.

Other listed node names are CLIPLoader, KSamplerAdvanced, LoraLoaderModelOnly, ModelSamplingSD3, PathchSageAttentionKJ, UNETLoader, VAEDecode, VAELoader, and VRAM_Debug.

Before setup, confirm your ComfyUI installation has matching components for CLIPLoader, KSamplerAdvanced, LoraLoaderModelOnly, VAEDecode, VAELoader, WanImageToVideo, CLIPTextEncode, PreviewImage, LoadImage, UNETLoader, ImageScale, and CreateVideo, and confirm the exact umt5_xxl ח

Sugestão

A license is not listed; check the source terms before using the workflow.

Sugestão