Doodle In LoRA for Qwen-Image 2.1 edits photos in 6 steps while keeping art style

Gradio · x · 2026-09-29

Gradio's "Doodle In" LoRA for Qwen-Image-2.1 lets you doodle anything into a photo while preserving the art style, offering editing flexibility and control that prompts alone struggle to deliver.

Edits finish in 6 steps using Viggle. Weights, training data and scripts are fully open on Hugging Face, with a ZeroGPU demo to try. Non-commercial research license.

Related event: Gradio Open-Sources Qwen-Image 2.1 Doodle-In LoRA(3 posts)→

Original post →

More from Multimodal

Multimodal channel →