This is an NF4 quantized model of Qwen image edit 2511 so it can run on GPUs using less than 20GB VRAM. You can run it on lower VRAM like 16GB. There were other NF4 models but they made the mistake of blindly quantizing all layers in the transformer. This one does not. We retain some layers at full precision in order to ensure that we get quality output. You can use the original Qwen Image Edit parameters. Model tested: Working perfectly even with 10 steps. Contact: support@JustLab.ai for commercial support, modifications and licensing. sample code FAQ 1. Does it support comfyui? Someone else has done an integration here https://github.com/mengqin/ComfyUI UnetBnbModelLoader 2. Is there a GGUF etc ? Not here. The motivation for this is to use Diffusers and get the best performance without relying on 3rd party toolkits and transformation wich break the pipeline. 3. How to use speed up lora? See 1 above. I believe the same lora(s) should work in the standard pipeline. In the past and aroudn the first release of Qwen Image a bug in diffusers prevented lora from attaching. 4. Can you support model XYZ? I only work with open source models that use a standard OSS licensing. Any custom lic…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy