Stable Diffusion 3.5 Large launches three new free ControlNet models: Blur, Canny & Depth

Stable Diffusion 3.5 Large has added new capabilities with the release of three ControlNets: Blur, Canny, and Depth, enhancing precision and usability in image generation for creative fields like interior design and architectural rendering.

The Blur ControlNet enables high-fidelity upscaling, suitable for converting low-resolution images into detailed visuals. Canny ControlNet uses edge maps to guide image structures, ideal for illustrations in various artistic styles. Depth ControlNet utilizes depth maps for precise control in architectural renderings and 3D texturing.

These models, available for free under the Stability AI Community License, allow both commercial and non-commercial use. Users can access the model weights on Hugging Face and the code on GitHub, with support for Comfy UI. Future updates will include Stable Diffusion 3.5 Medium variants and additional control types.