Originally posted to: <a target="_blank" rel="ugc" href="https://huggingface.co/Qwen/Qwen-Image-2512">https://huggingface.co/Qwen/Qwen-Image-2512</a>GGUF available here: <a target="_blank" rel="ugc" href="https://civarchive.com/models/2267491">https://civarchive.com/models/2267491</a>We are excited to introduce Qwen-Image-2512, the December update of Qwen-Image’s text-to-image foundational model. You are welcome to try the latest model at Qwen Chat. Compared to the base Qwen-Image model released in August, Qwen-Image-2512 features the following key improvements:<ul><li>Enhanced Human Realism Qwen-Image-2512 significantly reduces the “AI-generated” look and substantially enhances overall image realism, especially for human subjects.</li><li>Finer Natural Detail Qwen-Image-2512 delivers notably more detailed rendering of landscapes, animal fur, and other natural elements.</li><li>Improved Text Rendering Qwen-Image-2512 improves the accuracy and quality of textual elements, achieving better layout and more faithful multimodal (text + image) composition.</li></ul>We conducted over 10,000 rounds of blind model evaluations on <a target="_blank" rel="ugc" href="https://aiarena.alibaba-inc.com/corpora/arena/leaderboard?arenaType=T2I">AI Arena</a>, and the results show that Qwen-Image-2512 is currently the strongest open-source model—while remaining highly competitive even among closed-source models.<a target="_blank" rel="ugc" href="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/image2512/arena.png#center"><img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/image2512/arena.png#center" alt="" /></a>

Qwen-Image-2512

qwenImage2512_imageEdit2511.safetensors

Image Edit 2511

4 Step Turbo-LoRA

bf16

fp8_e4m3fn

The 2509 Workflow and LightX2v 8 Steps LoRA are compatible with this version.LoRA: <a rel="ugc" href="https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning/resolve/main/Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16.safetensors">Qwen-Image-Edit-2511-Lightning-4steps-V1.0-bf16</a> | 850MBLoRA: <a rel="ugc" href="https://huggingface.co/lightx2v/Qwen-Image-Edit-2511-Lightning/resolve/main/Qwen-Image-Edit-2511-Lightning-4steps-V1.0-fp32.safetensors">Qwen-Image-Edit-2511-Lightning-4steps-V1.0-fp32</a> | 1.7GBBF16 on 10GB VRAM + 64GB RAM: Workflow in the <a target="_blank" rel="ugc" href="https://civarchive.com/api/download/models/2530417?type=Training%20Data">zip file</a>. First, install SageAttention and Triton for your Python environment.( look for compatible precompiled wheels )<a target="_blank" rel="ugc" href="https://chat.qwen.ai/?inputFeature=image_edit">Online Demo</a> | <a target="_blank" rel="ugc" href="https://huggingface.co/Qwen/Qwen-Image-Edit-2511">HF</a> | <a target="_blank" rel="ugc" href="https://github.com/QwenLM/Qwen-Image">GH</a><h3 id="improvements-over-previous-versions-(2509)">Improvements Over Previous Versions (2509)</h3><ul><li>Significantly better overall consistency in edits, addressing drift issues.</li><li>Expanded multi-image capabilities.</li><li>Native LoRA integrations for enhanced effects like realistic lighting and viewpoint shifts.</li><li>Added support for industrial design and geometric reasoning, broadening use cases beyond basic editing.</li></ul><h3 id="examples">Examples</h3><ul><li>Character Consistency: Edit a portrait to change clothing or pose while keeping the face identical.</li><li>Multi-Person Fusion: Merge two individual photos into a group shot in a shared scene.</li><li>Lighting Enhancement: Apply realistic lighting using the integrated LoRA.</li><li>Viewpoint Generation: Create new angles of a subject without retraining.</li><li>Industrial Design: Batch-generate product variations and swap materials (e.g., change metal to wood).</li></ul>The model is mirrored here for convenience.

Qwen-Image-Edit 2511

qwenImageEdit2511_bf16_trainingData.zip

qwenImageEdit2511_bf16.safetensors

BF16

Files with SHA256: ae42d927b5fac4f278b9a894554c727e619727a63622976f2d95625be4bce08c