Wan 2.2 5B Image-to-Video Simple ComfyUI Workflow

This ComfyUI workflow is for image-to-video and text-to-video generation with Wan Video 2.2 TI2V-5B. It uses an image or text input to create a video output.

The 13-node workflow covers text encoding, image loading, Wan image-to-video latent creation, sampling, decoding, and video combination.

VHS_LoadImagePath handles image input, CLIPTextEncode handles text conditioning, and VHS_VideoCombine handles video output.

An accompanying video is listed as the usage reference.

The workflow is published on Civitai by lordcaocao2025.

The required model files include Wan2.2_VAE.pth and lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank128_bf16.safetensors.

The other required model files are wan2.2_ti2v_5B_fp16.safetensors and umt5_xxl_fp8_e4m3fn_scaled.safetensors.

It also lists comfyui-videohelpersuite, pinned to 08e8df15db24da292d4b7f943c460dc2ab442b24.

When configuring the workflow, review the KSamplerAdvanced, LoraLoader, ModelSamplingSD3, UNETLoader, and VAELoader nodes.

Suggestion · not verified

Before use, verify that all four model files are present with their exact filenames.

Suggestion · not verified

Also verify that comfyui-videohelpersuite is available at the listed pin.

Suggestion · not verified