Wan 2.2 14B Image to Video (Multiple Lora) Beginner Friendly Workflow Template — v2.0

Use this ComfyUI workflow to turn an image into video with Wan Video 14B i2v 720p, using separate high-noise and low-noise LoRAs as additional inputs.

The workflow follows the Wan 2.2 I2V template and replaces regular LoRA loaders with Power LoRA loaders from rgthree-comfy.

Provide an image, a custom high-noise LoRA, and a custom low-noise LoRA; the intended output is video.

Download and extract the JSON, place it in ComfyUI>User>default>workflows, then load “Wan 2.2 14B Multiple lora workflow Template.json.”

Vorschlag · nicht geprüft

Install rgthree-comfy through ComfyUI’s Custom Nodes Manager and restart ComfyUI before loading the workflow.

Vorschlag · nicht geprüft

The listed model files are umt5_xxl_fp8_e4m3fn_scaled.safetensors, wan_2.1_vae.safetensors, wan2.2_i2v_high_noise_14B_fp8_scaled.safetensors, and wan2.2_i2v_low_noise_14B_fp8_scaled.safetensors.

The workflow also requires the rgthree-comfy node pack.

Use the top Power LoRA Loader for high-noise LoRAs and the bottom loader for low-noise LoRAs; use +Add Lora to combine LoRAs.

Vorschlag · nicht geprüft

Pair wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors with a chosen custom high-noise LoRA, and pair wan2.2_i2v_lightx2v_4steps_lora_v1_low_noise.safetensors with a chosen custom low-noise LoRA.

Vorschlag · nicht geprüft

The instructions do not say whether a single-file LoRA works, so verify that case before relying on it.

Vorschlag

No license is specified in the listing; check the applicable terms before use.

Vorschlag