Awesome Stable Diffusion

A curated list of high-quality tools, UIs, libraries, and resources for Stable Diffusion and modern diffusion image models.

Contents

• Start here
• UIs
• Libraries
• Training
• Control and conditioning
• Upscale and restore
• Local and edge
• Integrations
• Video and 3D
• Desktop apps
• Contributing

Start here

• ComfyUI - Modular node-based GUI and backend for diffusion models.
• Diffusers - State-of-the-art diffusion models library for image, video, and audio in PyTorch.
• kohya_ss - GUI for training LoRA and DreamBooth models.

UIs

• ComfyUI - Most powerful modular diffusion GUI with a graph/nodes interface.
• Stable Diffusion web UI - Widely used Gradio-based Stable Diffusion web UI.
• automatic - Opinionated, actively maintained fork of Stable Diffusion WebUI.
• Fooocus - Minimal, user-friendly interface focused on good defaults.
• InvokeAI - Creative engine aimed at professionals and enthusiasts.
• ComfyUI-Manager - Install and manage custom nodes for ComfyUI.
• SwarmUI - Modular high-performance web UI for image and video diffusion models.
• stable-diffusion-webui-forge - Optimized A1111-based WebUI with improved VRAM management and speed.
• sd-webui-forge-classic - Actively maintained Forge Neo fork with FLUX, Wan, Qwen and modern model support.
• ComfyUI-KJNodes - Quality-of-life utility and model optimization nodes for ComfyUI.
• rgthree-comfy - Utility nodes and workflow improvements for ComfyUI.
• ComfyUI_LayerStyle - Photoshop-style layer, mask and compositing nodes for ComfyUI.
• ComfyUI-Custom-Scripts - Utility and workflow enhancement nodes for ComfyUI.
• ComfyUI-Easy-Use - Efficiency nodes that integrate common utilities for faster ComfyUI workflows.

Libraries

• Diffusers - Official Hugging Face library for diffusion pipelines and training helpers.
• LyCORIS - Extended LoRA-style adapters beyond conventional methods.
• flux - Official inference code for FLUX.1 open-weight models.
• DiffSynth-Studio - Unified diffusion engine with VRAM management, training and inference for FLUX, Wan and SD models.

Training

• sd-scripts - Core training scripts for DreamBooth and LoRA.
• kohya_ss - Popular GUI wrapper around kohya training scripts.
• LyCORIS - Alternative network architectures for efficient fine-tuning.
• ai-toolkit - All-in-one training toolkit for FLUX/SDXL and other diffusion models.
• OneTrainer - Unified GUI and scripts for training diffusion models including FLUX and SDXL.
• musubi-tuner - Kohya training scripts specialized for video diffusion models.

Control and conditioning

• ControlNet - Add spatial conditions (pose, depth, edges) to diffusion models.
• sd-webui-controlnet - ControlNet extension for AUTOMATIC1111 WebUI.
• sd-webui-additional-networks - Load LoRA and extra networks in A1111 WebUI.
• adetailer - Automatic detection and detailing extension for WebUI.
• ComfyUI_IPAdapter_plus - Reference IP-Adapter nodes for image prompt conditioning in ComfyUI.
• comfyui_controlnet_aux - Auxiliary preprocessors for ControlNet hint images in ComfyUI.
• x-flux-comfyui - ComfyUI nodes for FLUX ControlNets and LoRAs.
• ComfyUI-Advanced-ControlNet - Advanced ControlNet nodes with timestep scheduling and attention masks.
• ComfyUI-Fluxtapoz - ComfyUI nodes for Flux image editing via RF-Inversion and related methods.
• ComfyUI-BrushNet - Native BrushNet and PowerPaint inpainting nodes for any checkpoint.

Upscale and restore

• Real-ESRGAN - Practical AI image upscaler commonly used with Stable Diffusion.
• IOPaint - Inpainting and object removal powered by modern AI models.
• ComfyUI-Impact-Pack - Detector, detailer, and upscaler nodes for image enhancement in ComfyUI.

Local and edge

• stable-diffusion.cpp - Stable Diffusion and Flux inference in pure C/C++.
• MochiDiffusion - Native Stable Diffusion on Mac.
• web-stable-diffusion - Run Stable Diffusion in the browser with WebGPU.
• ComfyUI-GGUF - GGUF quantization support for native ComfyUI models (FLUX/SD3).
• mflux - Native MLX port of FLUX and related models for Apple Silicon.
• Mold - CLI-native local AI image/video/3D generation in Rust/Candle with CUDA/Metal and MCP.
• ComfyUI-nunchaku - ComfyUI nodes for 4-bit SVDQuant inference of FLUX and related models.

Integrations

• Krita AI Diffusion - Generate and inpaint with AI inside Krita.
• Dream textures - Stable Diffusion integration for Blender.
• Auto-Photoshop-StableDiffusion-Plugin - Generate images inside Photoshop via A1111 backend.

Video and 3D

• stable-diffusion-videos - Create videos by morphing between text prompts in latent space.
• i2vgen-xl - Video generation ecosystem built on diffusion models.
• Stable-Dreamfusion - Text-to-3D generation powered by Stable Diffusion.
• ComfyUI-WanVideoWrapper - Wrapper nodes for Wan video generation models in ComfyUI.
• ComfyUI-VideoHelperSuite - Nodes for video loading, combining, batching, and frame utilities in ComfyUI.
• ComfyUI-Frame-Interpolation - Video frame interpolation nodes supporting RIFE, FILM, GMFSS and other VFI models.
• ComfyUI-3D-Pack - Nodes for 3D mesh, 3DGS, and NeRF generation from images in ComfyUI.

Desktop apps

• Easy Diffusion - Simple one-click local installation for non-technical users.
• Off Grid AI Desktop - Local-first macOS app for on-device image generation.
• Imference Desktop - Local desktop app supporting SDXL, SD 1.5, FLUX, and custom weights.
• Image MetaHub - Search and organize AI images by prompt, model, LoRA, and workflow metadata.
• ImageBench - Open benchmark ranking text-to-image models with published outputs.
• StabilityMatrix - Multi-platform package manager and launcher for ComfyUI, Forge, A1111 and training tools.

Contributing

Contributions welcome. Please:

• Add only actively maintained, high-quality projects.
• Use the format: - [Name](url) - Short neutral description.
• Place items in the most relevant section.
• Open a pull request with a clear reason for the addition.

See the Awesome manifesto for curation guidelines.

License

To the extent possible under law, the contributors have waived all copyright and related or neighboring rights to this work.