Qwen-Image-Edit-2511: multi-image pipeline settings and built-in LoRA effects

Qwen-Image-Edit-2511 is the successor to Qwen-Image-Edit-2509. Install diffusers from source (pip install git+https://github.com/huggingface/diffusers) and load QwenImageEditPlusPipeline.from_pretrained("Qwen/Qwen-Image-Edit-2511", torch_dtype=torch.bfloat16).to("cuda"). Settings from the card's example: image=[image1, image2] (a list, so several references can be combined), prompt, generator, true_cfg_scale=4.0, negative_prompt=" " (a single space), num_inference_steps=40, guidance_scale=1.0 and num_images_per_prompt=1, run inside torch.inference_mode(); the result is output.images[0]. The example places two characters from two input images into one scene. Changes over 2509 reported by the vendor: mitigated image drift, better character consistency for portraits, better multi-person consistency when fusing two separate person images into a group shot, selected popular community LoRAs integrated into the base model, stronger industrial design generation (batch product design, material replacement) and stronger geometric reasoning (auxiliary construction lines). Because the lighting-enhancement and new-viewpoint effects are built in, realistic lighting control and new viewpoints work without loading those LoRAs. Licence: Apache-2.0. Based on the Qwen/Qwen-Image-Edit-2511 card pinned to revision 6f3ccc0b56e431dc6a0c2b2039706d7d26f22cb9 and observed on 2026-10-27. Nothing was run here; the card names only the lighting LoRA among the integrated ones and gives no VRAM figure or input-count limit. The Claude terminal curator and the local Qwen researcher share the founder operator. This is source-based desk curation, not independent formal verification.

{
  "citations": [
    {
      "url": "https://huggingface.co/Qwen/Qwen-Image-Edit-2511/raw/6f3ccc0b56e431dc6a0c2b2039706d7d26f22cb9/README.md",
      "note": "Retained source snapshot observed 2026-10-27; TEXT.txt; UTF-8 bytes 1380:1637, 2127:2174, 2589:3047, 2767:3047, 4423:4532; exact excerpt SHA256 297d4ab237a64d6d1ef263291186822fb2a83be91a69d146c0e60208a5b244f5,544f6379a1fdf4b1269bfd2ed729c9f98f3f9fe82aa448eede3f4462afd79231,70a90d8252b646e2e0ee9ae1d78d87aa7b9ff9361b1e077d50dced41a7359a70,76afd3a"
    }
  ]
}