Stable Diffusion WebUI Forge
Local AI image generation UI built on Stable Diffusion WebUI, with Flux, GGUF and LoRA support. Open source under AGPL-3.0.
Stable Diffusion WebUI Forge runs image generation on your own hardware through a browser interface. It builds on Stable Diffusion WebUI and suits people who want its image creation tools with more control over GPU memory use, as well as developers extending those tools. It's open source under AGPL-3.0.
Forge focuses on memory management and inference speed. Its GPU memory controls and model offloading let you adjust how much of a model stays on the GPU. The supplied Windows package uses PyTorch and CUDA for NVIDIA GPUs. Flux support includes full models and lower-bit formats such as BitsAndBytes NF4 and GGUF, with LoRA support for supported quantized variants.
The interface covers text-to-image and image-to-image generation, plus image upscaling. Integrated ControlNet tools, preprocessors and IP-Adapters give users additional ways to guide generation with image inputs. Instant-ID and reference-only methods are also supported. LayerDiffuse handles transparent image editing, while the canvas supports pressure input from Wacom tablets.
For people moving from Stable Diffusion WebUI, Forge can reuse existing checkpoints and extensions. It supports extensions such as Adetailer and includes FreeU. The Gradio interface provides checkpoint and LoRA selection, and text-to-image and image-to-image API endpoints let other software request generations.
