ComfyUI Workflows: Stable Diffusion, Flux, SDXL, ControlNet, image generation, AI art, and custom node-based diffusion pipelines.
๐ AWESOME COMFYUI BEYOND
The Ultimate 2024-2025 ComfyUI Resource Collection
From cutting-edge video generation to 3D creation, this is your comprehensive guide to the ComfyUI ecosystem.
๐ฌ Video Gen โข ๐ญ Face & Portrait โข ๐จ Image Magic โข ๐๏ธ 3D Creation โข โก Setup
Maintained by: umitkacar | Last Update: November 2024
๐ Table of Contents
๐ Click to expand navigation
- ๐ What's Hot in 2024-2025
- โก Quick Start & Essential Setup
- ๐ฌ Video Generation Revolution
- ๐ญ Face & Portrait Mastery
- ๐จ Image Enhancement & Effects
- ๐๏ธ 3D Generation & Depth
- ๐ง Workflow & Productivity
- ๐ฆ Latest Models (2024-2025)
- ๐ Learning Hub
- โญ Community Favorites
๐ What's Hot in 2024-2025
๐ฅ Major Breakthroughs
| Category | Innovation | Status | Why It Matters | | -------------- | ------------------------------ | ------------------------------------------------------------------------ | ------------------------------------------- | | ๐ฌ Video | HunyuanVideo (Tencent) | | 13B params, 5s videos, 8GB VRAM | | ๐ฌ Video | LTX Video (Lightricks) |
| First real-time DiT (faster than playback!) | | ๐จ Image | Flux.1 (Black Forest Labs) |
| Revolutionary quality, day-1 support | | ๐ญ Face | PuLID Flux II |
| Consistent faces, no model pollution | | ๐๏ธ 3D | Trellis |
| Unified 3D: Meshes, Gaussians, NeRF | | ๐ผ๏ธ Removal | BiRefNet |
| Best background removal (CAAI AIR'24) | | ๐ Depth | DepthCrafter |
| Consistent long depth sequences |
โ๏ธ ComfyUI V1 Desktop (October 2024)
๐ฆ One-Click InstallNo Python setup needed! |
๐ Model LibraryDrag & drop models |
๐ Auto UpdatesAlways stay current |
๐งฉ 600+ NodesCustom Node Registry |
โฌ๏ธ Download ComfyUI Desktop
โก Quick Start & Essential Setup
๐ฏ First-Time Setup Guide
graph LR
A[๐ฅ Download Desktop] --> B[๐ง Install Manager]
B --> C[๐ฆ Add Core Nodes]
C --> D[๐ฌ Try Workflow]
D --> E[๐ Create Magic!]
style A fill:#ff6b6b style B fill:#4ecdc4 style C fill:#45b7d1 style D fill:#96ceb4 style E fill:#ffeaa7
๐ ๏ธ Must-Have Extensions
๐ด CRITICAL - Install First! (Click to collapse)
1๏ธโฃ ComfyUI Manager 
https://github.com/Comfy-Org/ComfyUI-Manager
โ Install/update all custom nodes โ Automatic dependency resolution โ Model browser & downloader โ One-click missing node installation
2๏ธโฃ Workspace Manager 
https://github.com/11cafe/comfyui-workspace-manager
โ Git-like version control for workflows โ Organize workflows, models & images โ Sync across multiple devices โ Workflow history & rollback
๐จ Core Custom Nodes
| ๐ผ๏ธ Image Processing | โก Workflow Enhancement | ๐๏ธ Control & Preprocessing |
|---|---|---|
|
Impact Pack
โข Face detailing โข Segmentation โข Regional prompting IPAdapter Plus
โข 37 specialized nodes โข FaceID V2 support โข Style transfer ComfyUI Essentials
โข Missing core features โข Quality-of-life fixes |
WAS Node Suite
โข 100+ essential nodes โข Image processing โข Prompt utilities Efficiency Nodes
โข Streamlined workflows โข Reduced clutter rgthree-comfy
โข Power Prompt โข Fast Groups โข Utilities Custom Scripts
โข UI/UX improvements |
ControlNet Aux
โข 65+ preprocessors โข All control types Portrait Master
โข Advanced portrait control โข Professional results |
๐ฌ Video Generation Revolution
๐ State-of-the-Art Models (2024-2025)
๐ฅ Top Performers Comparison
| Model | Best For | Params | VRAM | Speed | Quality | | ------------------- | ----------------------- | ------ | ----- | ---------- | ---------- | | ๐ฅ HunyuanVideo | Professional Production | 13B | 8GB+ | โญโญโญ | โญโญโญโญโญ | | ๐ฅ LTX Video | Real-time Generation | 13B | 16GB+ | โญโญโญโญโญ | โญโญโญโญ | | ๐ฅ Mochi | Text-to-Video Quality | 10B | 20GB | โญโญ | โญโญโญโญโญ | | ๐ฏ CogVideoX | Image-to-Video | - | 14GB | โญโญโญ | โญโญโญโญ |
๐ด HunyuanVideo (Tencent) - CVPR 2025 Highlight
Repository: https://github.com/kijai/ComfyUI-HunyuanVideoWrapper
๐ฏ Key Features
- โจ 13B parameters, DiT architecture
- โจ Generate 5-second videos (extendable)
- โจ Native ComfyUI support (December 2024)
- โจ Custom 3D VAE compression
โ๏ธ Requirements
- Minimum VRAM: 8GB
- Recommended: 16GB+
- Best Quality: 24GB+
๐ฌ Use Cases
โ Professional video production
โ Marketing content
โ Social media videos
โ Product demonstrations
๐ LTX Video (Lightricks) - Real-Time Generation
Repository: https://github.com/Lightricks/ComfyUI-LTXVideo
๐ฏ Key Features
- ๐ FIRST real-time DiT video model
- ๐ Generate at 24FPS (768x512)
- ๐ Faster than playback speed!
- ๐ Supports up to 60 seconds
๐ฆ Model Variants
| Model | Steps | Speed | Quality | | ------------ | -------- | ------ | -------- | | Standard 13B | Normal | Fast | โญโญโญโญ | | Distilled | 4-8 only | โกโกโก | โญโญโญโญ |
โ๏ธ Requirements
- VRAM: 16GB+
- Quantized versions available for 4090/5090
๐ก Mochi (Genmo) - Best Text-to-Video Quality
Repository:
- https://github.com/kijai/ComfyUI-MochiWrapper
- https://github.com/logtd/ComfyUI-MochiEdit (Video editing)
๐ฏ Key Features
- ๐จ 10B parameters
- ๐จ Best text-to-video quality
- ๐จ Video editing capabilities
- ๐จ AsymmDiT architecture
โ๏ธ Requirements
- VRAM: 20GB (minimum)
- Originally: 4x H100 (100GB) - huge optimization!
๐ต CogVideoX (Tsinghua University) - Best I2V
Repository: https://github.com/kijai/ComfyUI-CogVideoXWrapper
๐ฏ Key Features
- ๐ธ Best image-to-video quality
- ๐ธ Chinese prompt optimization
- ๐ธ 3D VAE + Expert Transformer
- ๐ธ Efficient memory usage
โ๏ธ VRAM Usage
- Sampling: 5-6GB
- VAE Decode: 13-14GB (peak)
- With offloading: Works on 8GB cards
๐ ๏ธ Video Tools & Utilities
| Tool | Purpose | Stars | | --------------------------------------------------------------------------------- | ----------------------------- | -------------------------------------------------------------------------------------------------- | | VideoHelperSuite | Essential video processing | | | AnimateDiff-Evolved | Advanced AnimateDiff features |
| | LivePortraitKJ | Live portrait animation |
|
๐ญ Face & Portrait Mastery
๐ฏ Face Swapping Technology Comparison
graph TD
A[Choose Your Method] --> B{Priority?}
B -->|Speed| C[PuLID]
B -->|Quality| D[InstantID]
B -->|Flexibility| E[IPAdapter FaceID]
C --> F[โก <10 seconds] D --> G[๐ Best Results] E --> H[๐ง Most Versatile]
style C fill:#90EE90 style D fill:#FFD700 style E fill:#87CEEB
| ๐ฅ PuLID | ๐ฅ InstantID | ๐ฅ IPAdapter FaceID V2 |
|---|---|---|
|
Best for: Lightning-fast results
โจ Features:
|
Best for: Maximum quality
โจ Features:
|
Best for: Integration needs
โจ Features:
|
๐ Face Enhancement Tools
| Tool | Purpose | Key Feature | | ----------------------------------------------------------------- | ------------ | ------------------------- | | ๐จ Impact Pack | FaceDetailer | Automatic face refinement | | ๐ ReActor | Face Swap | Easy face swapping | | ๐ญ FaceFusion | Standalone | Professional face fusion |
๐จ Image Enhancement & Effects
๐ Best-in-Class Tools (2024)
๐ผ๏ธ Background Removal - BiRefNet
|
Original Implementation:
Multi-Model Integration:
|
๐ฏ Why BiRefNet?
|
๐ฆ Model Variants
| Model | Resolution | Use Case | | -------- | ---------- | --------------- | | General | Standard | Everyday use | | HR | 2048x2048 | High resolution | | Lite | Fast | Speed priority | | Portrait | Optimized | Human portraits |
๐ Upscaling Solutions
| ๐ฅ SUPIR - Professional Grade | ๐ฅ AuraSR - GAN Power |
|---|---|
๐ Features:
|
๐ Features:
โญ Rating:
|
๐ก IC-Light - Professional Relighting
|
Implementations:
|
๐ฌ Capabilities:
|
๐ Licensing
| Version | License | Commercial Use | | ------- | ---------- | ---------------------- | | V1 | Open | โ Allowed | | V2 | Restricted | โ Non-commercial only |
๐จ Advanced Inpainting
PowerPaint V2 + BrushNet (2024)
Repository: https://github.com/nullquant/ComfyUI-BrushNet
๐ฏ Features
- ๐๏ธ BrushNet + PowerPaint integration
- ๐๏ธ Object removal with SAM
- ๐๏ธ Convert any SD 1.5 to inpainting model
- ๐๏ธ Multiple inpainting modes
ControlNet Union - Unified Control (2024)
For Flux
github.com/XLabs-AI/x-flux-comfyui
7 Control Modes: canny, tile, depth, blur, pose, gray, low quality
For SDXL
Includes ALL control types in ONE model!
- โ Massive storage savings
- โ Promax: +Tile Variation, Deblur, Super-resolution
๐ Depth Estimation
DepthCrafter
Repository: https://github.com/akatz-ai/ComfyUI-DepthCrafter-Nodes
๐ฏ Revolutionary Features:
- ๐ฌ Consistent long depth sequences
- ๐ฌ No optical flow needed
- ๐ฌ No camera pose required
- ๐ฌ Video editing & post-production
- Window size: 75-110 (recommended)
- Overlap for longer videos
- CPU offload saves +25% VRAM
๐๏ธ 3D Generation & Depth
Trellis - Unified 3D Asset Generation
|
Repositories:
|
๐ฏ Output Formats:
|
๐ฌ Technology
SLAT Representation (Structured 3D Latents)
- Unified format for all 3D types
- Editable 3D assets
- Scalable & versatile
โ๏ธ Requirements
- Minimum VRAM: 8GB
- Recommended: 16GB+
๐ ๏ธ Additional 3D Tools
| Tool | Purpose | | ------------------------------------------------------------------ | ---------------------------------------- | | ComfyUI-3D-Pack | Comprehensive 3D processing (3DGS, NeRF) | | Apple ml-4m | Multimodal model research |
๐ง Workflow & Productivity Tools
โก Essential Utilities
| ๐ฅ๏ธ Monitoring | โ๏ธ Optimization | โ๏ธ Prompting |
|---|---|---|
|
Crystools
Real-time monitoring:
|
Use Everywhere
|
SDXL Prompt Styler
|
๐ฆ Latest Models (2024-2025)
๐ฅ Flux.1 (Black Forest Labs) - August 2024
Built-in Nodes: FluxGuidance, ModelSamplingFlux
Flux.1 DevFeatures:
|
Flux.1 SchnellFeatures:
|
Extensions:
- x-flux-comfyui - LoRA & ControlNet integration
๐ Stable Diffusion 3 / 3.5 (2024)
| Model | Release | Download | | ---------------- | --------- | ----------------------------------------------------------------------------- | | SD3 Medium | June 2024 | ๐ฅ Download | | SD3.5 Large | Oct 2024 | ๐ฅ Download | | SD3.5 Medium | Oct 2024 | ๐ฅ Download |
๐ Kolors (Kuaishou) - July 2024
Repository: https://github.com/Kwai-Kolors/Kolors ComfyUI: https://github.com/MinusZoneAI/ComfyUI-Kolors-MZ
๐ Features:
- Bilingual excellence (Chinese & English)
- Trained on billions of text-image pairs
- KlingAI project creators
๐ฆ Classic Models (SDXL, SD 1.5, SD 2.x) - Click to expand
SDXL Models
| Model | Download | | ------------------------ | ------------------------------------------------------------------------------------------------------------------- | | SDXL Base | ๐ฅ | | SDXL Refiner | ๐ฅ | | DreamShaper Lightning XL | ๐ฅ | | SDXL Turbo | ๐ฅ |
SD 1.5 & 2.x Models
| Model | Download | | ------------ | --------------------------------------------------------------------------------------------------------------- | | SD v1.5 | ๐ฅ | | SD 2.1 (512) | ๐ฅ | | SD 2.1 (768) | ๐ฅ |
๐ Learning Hub
๐ Official Documentation
๐ Docs2024 Updated |
๐ BlogLatest news |
๐ฏ ExamplesLearn by doing |
๐จ TemplatesReady to use |
๐ฌ Video Tutorials & Courses
| Resource | Type | Level | Cost | | -------------------------------------------------------------------- | -------- | ----------------- | ---- | | Olivio Sarikas | YouTube | All levels | Free | | ThinkDiffusion | Platform | Beginner-Advanced | Free | | FastCampus | Course | Comprehensive | $215 |
๐ Community Platforms
| Platform | Workflows | Features | | ----------------------------------------------------------- | ----------- | ------------------- | | OpenArt Academy | Educational | Structured learning | | OpenArt Templates | 1000+ | Searchable | | ComfyWorkflows | Community | Sharing platform | | RunComfy | 200+ | Curated |
โ๏ธ Cloud Platforms
| ๐ฅ๏ธ Comflowy | โก RunComfy |
|---|---|
|
Features:
|
Features:
|
โญ Community Favorites & Top Repos
๐ Most Starred (2024)
| Rank | Repository | Stars | Description | | ---- | ------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------- | --------------------------------------- | | ๐ฅ | ZHO's Workflows | | Comprehensive Chinese/English workflows | | ๐ฅ | HunyuanVideoWrapper |
| Latest video generation | | ๐ฅ | AdvancedLivePortrait |
| Advanced portrait animation |
๐ Essential Resources
| Resource | Purpose | | ---------------------------------------------------------------------- | --------------------------------- | | Top 100 ComfyUI | Auto-updated ranking of top repos | | Awesome ComfyUI | Curated node collection |
๐ญ Specialized Extensions
๐ค Vision & Language Models
- Florence-2 - Microsoft's vision model
- Florence2 Alt - Alternative implementation
- IFAI_tools - LLM integration
- VLM nodes - Vision Language Models
โ๏ธ Segmentation & Detection
- Segment Anything - SAM integration
- Depth Anything V2 - Latest depth estimation
๐ฌ Animation & Motion
- AnimateDiff - Animation generation
- DragDiffusion - Interactive editing
- MusePose - Pose-driven animation
- MimicMotion - Motion transfer
- Hallo - Portrait animation
๐ง Utilities & Integration
- Tooling Nodes - Professional tools
- Inpaint Nodes - Advanced inpainting
- API Tools - API integration
- Blender Integration - 3D workflow
- Krita Integration - Painting software
- Enhanced Sidebar - UI improvement
๐ค Contributing
Help Make This Resource Better!
How to contribute:
- ๐ด Fork this repository
- โ Add your resource to appropriate section
- ๐ Include brief, clear description
- โ Verify links work
- ๐ Create pull request
- โ ComfyUI-related only
- โ 2024-2025 content preferred
- โ Working links required
- โ Clear descriptions
- โ Proper categorization
๐ Important Notes
๐พ Model Downloads
Requirements may include:
- Hugging Face account + license agreement
- Civitai account
- Sufficient disk space (10-100GB per model)
- VRAM as specified
๐ Commercial Use Licensing
| โ Commercial Friendly | โ ๏ธ Non-Commercial Only |
|---|---|
|
|
๐ More Awesome Resources
- ComfyUI Official - The original project
- ComfyUI Manager - Official registry
- Awesome ComfyUI - Another great list
- ComfyUI Workflows - Workflow focused
๐ก Quick Start Checklist
New to ComfyUI? Follow this path:
โ
1. Download ComfyUI Desktop V1
โ
2. Install ComfyUI Manager
โ
3. Watch Olivio Sarikas tutorials
โ
4. Try official example workflows
โ
5. Explore OpenArt templates
โ
6. Join the community!
โญ Star History
If this helped you, please star this repo!
It helps others discover these resources.
๐ Repository Stats
This is a living document. ComfyUI evolves rapidly - this list is continuously updated with cutting-edge developments.
Questions? Ideas? Contributions? Open an issue or submit a pull request!
Maintained with โค๏ธ by umitkacar
Last Update: January 2025
๐ ๏ธ For Developers
This repository is built with modern Python development practices and comprehensive quality controls.
๐ฆ Repository Features
- โ 100% Test Coverage - Comprehensive test suite with pytest
- โ Type Safety - Full MyPy strict mode compliance
- โ Security Audited - pip-audit, bandit, and safety scanning
- โ Modern Tooling - Hatch, Ruff, Black, pre-commit hooks
- โ CI/CD Ready - Automated testing and quality checks
- โ Parallel Testing - pytest-xdist for fast test execution
๐ Quick Development Setup
# Clone repository
git clone https://github.com/umitkacar/Awesome-ComfyUI-Beyond.git
cd Awesome-ComfyUI-Beyond
Install with all development tools
pip install -e ".[dev]"
Or use Hatch (recommended)
pip install hatch
make install
Run all checks
make check-all
๐งช Available Commands
make test # Run tests
make test-cov # Run tests with coverage report
make test-parallel # Run tests in parallel (16 workers)
make lint # Run linting (Ruff)
make format # Format code (Black)
make type-check # Type checking (MyPy)
make audit # Dependency vulnerability audit
make security # Security scanning (Bandit)
make check-all # Run all quality checks
๐ Documentation
- DEVELOPMENT.md - Complete development environment setup
- CONTRIBUTING.md - Contribution guidelines
- CHANGELOG.md - Version history and changes
- LESSONS-LEARNED.md - Technical decisions and solutions
๐ Security
This repository undergoes comprehensive security auditing:
# Dependency vulnerability scanning
make audit
Python security code analysis
make security
All security checks pass โ
Security Status: ๐ข No known vulnerabilities
๐ Code Quality Metrics
| Metric | Score | Tool | | -------------- | -------- | ------------------ | | Test Coverage | 100% | pytest-cov | | Type Coverage | 100% | MyPy (strict) | | Security Score | 100% | bandit + pip-audit | | Code Quality | A+ | Ruff | | Documentation | Complete | Comprehensive docs |
๐ค Contributing
Contributions are welcome! Please see CONTRIBUTING.md for guidelines.
๐ License
MIT License - see LICENSE for details.