Custom Node:

https://github.com/lrzjason/Comfyui-QwenEditUtils

The V2 LoRA has been trained for QwenEditPlus (aka QwenImage Edit, 2509).

When processing multiple images, a strength value between 0.5 and 1.0 is recommended.

For a single image, the value range is 0–0.5.

Needs to be used with Workflow v4.0: https://civitai.com/models/1939540

Trained on qwenvl without reference images.

This is a typical method to test consistency with context (e.g., workflows).

Recommended LoRA strength: 0.45–0.5

Workflow:

https://civitai.com/models/1939540/qwenedit-consistance-edit-workflow