Qwen-Image
An open-source image generation and editing model that runs locally, renders Chinese text and accepts multiple images for edits.
Alternative to:
Qwen-Image generates images from text and edits existing pictures. It's for creators and developers who want to run an image model on their own hardware, especially for work that needs readable text within images.
- Text rendering and editing: Create images with multiline text and paragraph layouts, including Chinese lettering, or revise text in an existing picture.
- Realistic people: Generate people with detailed faces, textures and surrounding scenes.
- Portrait and group edits: Use single or multiple input images to change a scene while preserving a person's identity and visual traits, or combine separate portraits into a group picture.
- Lighting effects: The editing model includes selected community LoRA effects, including lighting control that doesn't require extra tuning.
Qwen-Image is open source under Apache 2.0 and works with Hugging Face Diffusers. Local deployment supports a multi-GPU API server with a Gradio browser interface, queued requests and multiple aspect ratios. Qwen Chat also offers hosted image generation.
Optional prompt enhancement uses Qwen-Plus for image generation and Qwen-VL-Max for editing. Editing results can be unstable without prompt rewriting.
Logotyper
