The documentation lists improvements including reduced image drift, character and multi-person consistency, text editing, material replacement, and geometric annotation use cases.
The workflow is available through ComfyUI’s Template Library under “Qwen-Image-Edit-2511,” as a downloadable JSON file, or through Comfy Cloud. For local use, the page specifies a Qwen 2.5 VL 7B FP8 text encoder, the Qwen Image Edit 2511 BF16 diffusion model, and a VAE, stored in separate ComfyUI model directories. An optional Lightning LoRA is listed for four-step acceleration.
ComfyUI says users should update their installation. It notes that missing core nodes can mean the required nodes have not yet reached the latest stable release and may require the Nightly version; Comfy Cloud will update after stable release.
The page does not establish the model’s licence, hardware requirements, benchmark results, or a date for stable availability.