ComfyUI

Main Features:

• Nodes/graph/flowchart interface to experiment and create complex Stable Diffusion workflows without needing to code anything.
• Image Models: SD1.x, SD2.x, SDXL, SDXL Turbo, Stable Cascade, SD3, SD3.5, Pixart Alpha/Sigma, AuraFlow, HunyuanDiT, Flux, Lumina Image 2.0, HiDream, Qwen Image, Hunyuan Image 2.1.
• Image Editing Models: Omnigen 2, Flux Kontext, HiDream E1.1, Qwen Image Edit.
• Video Models: Stable Video Diffusion, Mochi, LTX-Video, Hunyuan Video, Wan 2.1, Wan 2.2.
• Audio Models: Stable Audio, ACE Step.
• 3D Models: Hunyuan3D 2.0.
• Asynchronous Queue system.
• Can load ckpt and safetensors: All in one checkpoints or standalone diffusion models, VAEs and CLIP models. Safe loading of ckpt, pt, pth, etc.. files.
• Embeddings/Textual inversion, Loras (regular, locon and loha), Hypernetworks.
• Loading full workflows (with seeds) from generated PNG, WebP and FLAC files. Saving/Loading workflows as Json files.
• Area Composition, Inpainting with both regular and inpainting models, ControlNet and T2I-Adapter, Upscale Models (ESRGAN, etc.), GLIGEN, Model Merging, LCM models and Loras, Latent previews with TAESD.
• Works fully offline: core will never download anything unless you want it to.
• Optional API nodes to use paid models from external providers through the online Comfy API.
• Support for async node functions and execution.

Main Characteristics:

• Many optimizations: Only re-executes the parts of the workflow that changes between executions.
• Smart memory management: can automatically run large models on GPUs with as low as 1GB vram with smart offloading.
• Works even without a GPU with: --cpu (slow).
• Config file to set the search paths for models to share between another UI and ComfyUI.
• Supports TLS/SSL via self-signed certificates using command line arguments.

Instructions and Workflow:

1. Windows Portable: Download, extract, place checkpoints in ComfyUI\models\checkpoints, and run. Portable builds available for Nvidia and AMD GPUs.
2. Manual Install (Windows, Linux): Git clone, place models, install PyTorch and dependencies (pip install -r requirements.txt), run python main.py.
3. Apple Mac silicon: Install pytorch nightly, install dependencies, run python main.py.
4. Other Hardware: Supports AMD GPUs (ROCm), Intel GPUs (IPEX/XPU), Ascend NPUs, Cambricon MLUs, Iluvatar Corex with specific PyTorch extension installations.
5. Frontend Development: New frontend is default. Can use --front-end-version argument to use the latest daily or specific version, or access the legacy frontend.

Core Advantages: Highly modular, supports a vast range of models (image, video, audio, 3D), excellent smart memory and dependency-aware caching optimization, fully open-source and offline-capable, cross-platform support for diverse hardware (NVIDIA, AMD, Intel, Apple Silicon, Ascend, etc.).

Target Users: AI image/video/audio generation developers, advanced Stable Diffusion users, designers requiring complex workflows, users looking to run large models on low VRAM devices.

Pricing Information: The ComfyUI core software is completely free and open-source (GPL-3.0 license). It offers optional API nodes to access paid models from external providers via the online Comfy API, which involves external provider fees.