Perform Precise In-Image Text Editing with Qwen-Image-Edit
Edit text directly within images while maintaining stylistic consistency.
Use Qwen-Image-Edit to modify text within existing images, ensuring the new text matches the original font, layout, and visual environment.
The Scenario
You have a social media graphic or a presentation slide where a word is misspelled or a product name needs to be updated, but you don't have the original design project file.
Before & after
Editing text inside an existing image usually requires graphic design software like Photoshop, involving masking, font matching, and manual blending, taking 15–30 minutes.
Using Qwen-Image-Edit, you can simply type the text change you want; the model handles the rendering and semantic blending in under 1 minute.
The Prompt
Edit this image to change the text from '[OLD_TEXT]' to '[NEW_TEXT]'. Maintain the original font style, color, and background texture: [UPLOAD_IMAGE]
Qwen-Image-Edit uses visual semantic control (Qwen2.5-VL) and appearance control to ensure that new text matches the original image's style and lighting.
Source
Blog | Qwen"Qwen-Image-Edit successfully extends Qwen-Image’s unique text rendering capabilities to image editing tasks, enabling precise text editing."