🎨 DisComfy - Discord AI Art Bot

Version 2.2.0 | Advanced AI Image & Video Generation for Discord

The most advanced Discord bot for AI image and video generation! DisComfy seamlessly integrates with ComfyUI to bring professional-grade AI art generation directly to your Discord server. Create stunning visuals, upscale images, and generate videos with real-time progress tracking and interactive parameter selection.

🚀 Current Version: v2.2.0 📂 Repository: https://github.com/jmpijll/discomfy.git 🎯 Status: Production Ready

🎉 What's New - Qwen Image 2512 Model

🧠 NEW Model: Qwen Image 2512 - High-quality text-to-image with built-in hi-res fix
🔬 Hi-Res Fix Pipeline: 1.5x latent upscale with dedicated refinement pass
📐 Portrait Optimized: Default 1296x1728 (3:4 ratio)
✨ LoRA Support: Compatible with Qwen Image 2512 LoRAs
⚡ Fast Generation: Only 8 steps at CFG 0.8 for the initial pass
🔧 Workflow Enable/Disable: Control models via environment variables

📝 Previous Releases

v2.1.2 - ZI Turbo Model:

⚡ ZI Turbo ultra-fast turbo generation
🔥 Only 10 steps with CFG 1.0 for rapid generation
🛠️ KSamplerAdvanced workflow updater

v2.1.0 - Ultra High-Resolution Support:

🚀 DyPE Flux Krea model with 4K resolution support
📐 Dynamic Position Encoding technology
⚙️ Adjustable DyPE exponent control

v2.0.0 - Architectural Overhaul:

🏗️ Complete modular architecture (77% code size reduction)
✅ 85/86 tests passing (99% pass rate)
📚 24+ comprehensive documentation guides
🐳 Docker support with GHCR and Docker Hub

See RELEASE_NOTES.md for complete details.

🌟 Why Choose DisComfy?

DisComfy isn't just another AI bot - it's a complete creative powerhouse that brings the full capabilities of ComfyUI to Discord with an intuitive, user-friendly interface. Whether you're an artist, content creator, or AI enthusiast, DisComfy makes it incredibly easy to create stunning visuals right in your Discord server.

🔥 Key Highlights:

Real-time Progress Tracking with step-by-step updates
Interactive Parameter Selection for complete customization
Custom Animation Prompts - modify how your images animate
Advanced Image Editing with natural language prompts
Extended Video Generation with 15-minute timeout support
Professional-grade Quality using ComfyUI workflows
Community Friendly - anyone can use any generation's buttons
Zero Downtime - buttons never expire, unlimited usage

✨ Core Features

🎨 AI Image Generation

High-Quality Output: Generate stunning images using advanced ComfyUI workflows
Multiple Models: Support for Flux, Flux Krea, DyPE Flux Krea (4K), ZI Turbo, Qwen Image 2512 🧠 NEW, HiDream, and custom workflows
Hi-Res Fix Pipeline: Qwen Image 2512 features a built-in 1.5x latent upscale with refinement pass
Ultra High-Resolution: DyPE Flux Krea supports resolutions up to 4096x4096 with advanced position encoding
LoRA Integration: Dynamic LoRA loading with customizable strength
Flexible Parameters: Control width, height, steps, CFG, batch size, DyPE exponent, and seed
Batch Generation: Create multiple images in a single request

🔍 Advanced Image Upscaling

Multiple Ratios: Choose between 2x, 4x, or 8x magnification
AI Super-Resolution: Professional-grade upscaling using ComfyUI
Customizable Parameters: Adjust denoise strength and sampling steps
Interactive Selection: Discord modals for parameter customization
Original Prompt Preservation: Automatically uses original image prompts

🎬 Professional Video Generation

High-Quality Animation: Convert images to smooth MP4 videos
Custom Animation Prompts: Modify how your images animate with natural language
Pre-filled Defaults: Original prompts auto-filled, edit as needed
Multiple Frame Counts: Choose 81, 121, or 161 frames for different lengths
Extended Timeout: 15-minute timeout supports complex video workflows
Interactive Settings: Customize strength, steps, and frame count via modal
Real-time Progress: Track every step of video generation

✏️ Advanced Image Editing

Dual Editing Models: Choose between Flux Kontext or Qwen 2.5 VL for editing
Natural Language Editing: Describe changes in plain English
Flux Kontext: High-quality editing with 10-50 steps (slower, more detailed)
Qwen 2.5 VL: Ultra-fast editing with 4-20 steps (faster, efficient)
Multi-Image Qwen Editing: NEW in v1.3.0 - Edit with 1, 2, or 3 input images
Direct Upload Support: Edit any image with /editflux or /editqwen commands
Post-Generation Editing: Both edit buttons available on all generated images
Customizable Parameters: Adjust sampling steps for quality control

📊 Real-Time Progress Tracking

Step-Based Accuracy: Progress calculated from actual sampling steps, not time estimates
WebSocket Integration: Real-time updates directly from ComfyUI
Detailed Information: See current node, step progress, and time estimates
Queue Monitoring: Live updates when waiting in ComfyUI's queue
Fallback Support: Graceful degradation to HTTP polling when needed

⚙️ Interactive User Experience

Parameter Selection Modals: Customize settings before generation starts
Universal Button Access: Anyone can use action buttons on any generation
Infinite Usage: Buttons never expire and can be used multiple times
Smart Rate Limiting: Prevents abuse while allowing normal usage
Error Handling: Robust error recovery and user-friendly messages

🛡️ Production-Ready Features

Automatic File Management: Intelligent cleanup of old outputs
Concurrent Request Handling: Multiple users can generate simultaneously
Session Management: Proper HTTP session handling with connection pooling
Security: Input validation, rate limiting, and secure file handling
Comprehensive Logging: Detailed logs for debugging and monitoring

🚀 Real-Time Progress Tracking

DisComfy features advanced progress tracking that provides detailed information about your generation:

📈 Progress Display:

🎬 Generating Video
📊 87.5% ████████████████████░░░
🔄 Sampling (315/321)
⏱️ Elapsed: 4m 32s | ETA: 42s
🎯 Node: WanVaceToVideo
Settings: 720x720 | 161 frames | Strength: 0.7

🔧 Technical Features:

WebSocket Integration: Direct connection to ComfyUI for real-time updates
Step-Based Calculation: Accurate progress based on actual sampling steps
Node Execution Tracking: See which ComfyUI nodes are currently running
Cached Node Detection: Automatically accounts for skipped nodes
Multi-Phase Support: Handles complex workflows with multiple sampling stages
Automatic Fallback: HTTP polling when WebSocket unavailable

⚙️ Interactive Parameter Selection

Take full control of your generations with interactive Discord modals:

🔍 Upscale Customization:

Upscale Ratio: 2x, 4x, or 8x magnification
Denoise Strength: 0.1 - 1.0 for detail enhancement
Sampling Steps: 10-50 steps for quality control

🎬 Video Customization:

Frame Count: 81, 121, or 161 frames (2-5 second videos)
Animation Strength: 0.1 - 1.0 for intensity control
Quality Settings: Fine-tune your video output

🎯 Smart Defaults:

Pre-filled with sensible defaults
Input validation prevents errors
Real-time parameter descriptions
Original prompts automatically preserved

🚀 Quick Start Guide

Prerequisites:

Python 3.8+ installed
ComfyUI running locally or remotely
Discord Bot Token
Basic command line knowledge

1. Clone & Setup:

git clone https://github.com/jmpijll/discomfy.git
cd discomfy
python -m venv venv

# Activate virtual environment
source venv/bin/activate  # Linux/Mac
# or
venv\Scripts\activate     # Windows

pip install -r requirements.txt

2. Configure Bot:

cp config.example.json config.json
# Edit config.json with your Discord token and ComfyUI URL

Example config.json:

{
  "discord": {
    "token": "YOUR_DISCORD_BOT_TOKEN",
    "guild_id": "YOUR_SERVER_ID"
  },
  "comfyui": {
    "url": "http://localhost:8188",
    "timeout": 300
  },
  "generation": {
    "default_workflow": "flux_lora",
    "max_batch_size": 4,
    "output_limit": 100
  }
}

3. Discord Bot Setup:

Go to Discord Developer Portal
Create new application → Bot
Copy bot token to config.json
Invite bot with permissions: Send Messages, Use Slash Commands, Attach Files, Embed Links

4. Prepare ComfyUI:

Ensure ComfyUI is running: http://your-comfyui-url/system_stats
Place workflow JSON files in workflows/ folder
Install required models and LoRAs

5. Launch Bot:

python bot.py

Success Output:

🤖 Bot is starting up...
✅ Connected to Discord as DisComfy#0430
🎨 ComfyUI connection verified
🚀 Bot is ready! Use /generate to start creating!

🐳 Docker Quick Start

DisComfy is available as a pre-built Docker container from both GitHub Container Registry (ghcr.io) and Docker Hub.

Using the Pre-built Image:

From GitHub Container Registry:

# Pull the latest image
docker pull ghcr.io/jmpijll/discomfy:latest

# Or pull a specific version
docker pull ghcr.io/jmpijll/discomfy:v2.2.0

From Docker Hub:

# Pull the latest image
docker pull jamiehakker/discomfy:latest

# Or pull a specific version
docker pull jamiehakker/discomfy:v2.2.0

Running with Docker:

docker run -d \
  --name discomfy \
  -v $(pwd)/config.json:/app/config.json:ro \
  -v $(pwd)/workflows:/app/workflows:ro \
  -v $(pwd)/outputs:/app/outputs \
  -v $(pwd)/logs:/app/logs \
  -e DISCORD_TOKEN=your_discord_token \
  -e COMFYUI_URL=http://your-comfyui-url:8188 \
  ghcr.io/jmpijll/discomfy:latest

Docker Compose Example:

Create a docker-compose.yml:

version: '3.8'

services:
  discomfy:
    # Use either:
    # image: ghcr.io/jmpijll/discomfy:latest
    # or:
    image: jamiehakker/discomfy:latest
    container_name: discomfy
    restart: unless-stopped
    volumes:
      - ./config.json:/app/config.json:ro
      - ./workflows:/app/workflows:ro
      - ./outputs:/app/outputs
      - ./logs:/app/logs
    environment:
      - DISCORD_TOKEN=${DISCORD_TOKEN}
      - COMFYUI_URL=${COMFYUI_URL:-http://localhost:8188}
    env_file:
      - .env

Then run:

docker-compose up -d

Building from Source:

# Clone the repository
git clone https://github.com/jmpijll/discomfy.git
cd discomfy

# Build the Docker image
docker build -t discomfy:local .

# Run the container
docker run -d --name discomfy \
  -v $(pwd)/config.json:/app/config.json:ro \
  ghcr.io/jmpijll/discomfy:local

Note: Make sure your config.json and workflows/ directory are accessible to the container, and that outputs/ and logs/ directories exist for the container to write to.

📖 Usage Examples

Basic Image Generation:

/generate prompt:a majestic dragon soaring through clouds

Image Editing (Flux Kontext - High Quality):

/editflux image:photo.jpg prompt:add sunglasses and a hat steps:25

Image Editing (Qwen - Ultra Fast):

/editqwen image:photo.jpg prompt:add sunglasses and a hat steps:8

Multi-Image Qwen Editing (NEW in v1.3.0):

# Single image (default)
/editqwen image:photo.jpg prompt:enhance the lighting steps:8

# Two images
/editqwen image:photo1.jpg image2:photo2.jpg prompt:combine both styles steps:8

# Three images
/editqwen image:photo1.jpg image2:photo2.jpg image3:photo3.jpg prompt:merge all three steps:8

Advanced Parameters:

/generate prompt:cyberpunk cityscape at night width:1024 height:768 steps:30 cfg:7.5 batch_size:2

Using LoRAs:

/generate prompt:anime character in magical forest lora:anime_style_v2 lora_strength:0.8

Post-Generation Actions:

Generate an image with /generate
Click 🔍 Upscale → Choose 4x ratio, adjust settings
Click ✏️ Flux Edit → High-quality editing with Flux Kontext
Click ⚡ Qwen Edit → Ultra-fast editing with Qwen 2.5 VL
Click 🎬 Animate → Select 161 frames, set strength
Anyone can use these buttons on any generation!

🔧 Advanced Configuration

Custom Workflows:

Add your ComfyUI workflows to the workflows/ folder and configure them in config.json:

"workflows": {
  "my_custom_workflow": {
    "file": "my_workflow.json",
    "name": "Custom Style",
    "description": "My custom ComfyUI workflow",
    "enabled": true
  }
}

Enable/Disable Workflows:

Control which models and features are available via environment variables:

# In your .env file
WORKFLOW_HIDREAM_LORA_ENABLED=false
WORKFLOW_VIDEO_WAN_VACE_14B_I2V_ENABLED=false

All workflows are enabled by default. This is useful for:

Retiring old models
Reducing server load
Testing specific features
Maintenance mode

See Workflow Management Guide for complete details

Performance Tuning:

Image Generation: ~30 seconds average
Video Generation: 2-10 minutes (15-minute timeout)
Upscaling: ~45 seconds average
Concurrent Users: Fully supported
Memory Usage: Auto-cleanup after 100 files

🛠️ Troubleshooting

Common Issues:

Bot won't start:

Verify Discord token in config.json
Check bot permissions in Discord server
Ensure Python 3.8+ is installed

ComfyUI connection failed:

Verify ComfyUI is running: visit URL in browser
Check firewall settings
Ensure correct URL in config.json

Generation fails:

Check ComfyUI logs for errors
Verify required models are installed
Try simpler parameters first

Timeout errors:

Video generation can take up to 15 minutes
Check ComfyUI system resources
Consider reducing batch sizes

Getting Help:

Check logs in logs/bot.log
Verify ComfyUI system stats
Test with simple generations first
Join our Discord community for support

📁 Project Structure

discomfy/
├── bot.py                 # Main Discord bot
├── image_gen.py          # Image generation engine  
├── video_gen.py          # Video generation engine
├── config.py             # Configuration management
├── requirements.txt      # Python dependencies
├── config.json          # Bot configuration
├── workflows/           # ComfyUI workflow files
│   ├── flux_lora.json
│   ├── flux_krea_lora.json    # Enhanced Flux Krea model
│   ├── dype-flux-krea-lora.json # DyPE 4K model
│   ├── ZITURBO1.json          # ZI Turbo fast generation
│   ├── qwen_image_2512_lora.json # Qwen Image 2512 with hi-res fix
│   ├── flux_kontext_edit.json # Flux Kontext editing
│   ├── qwen_image_edit.json   # Qwen 2.5 VL fast editing (1 image)
│   ├── qwen_image_edit_2.json # Qwen 2.5 VL multi-image (2 images)
│   ├── qwen_image_edit_3.json # Qwen 2.5 VL multi-image (3 images)
│   ├── hidream_lora.json
│   ├── upscale_config-1.json
│   └── video_wan_vace_14B_i2v.json
├── outputs/             # Generated files
├── logs/               # Bot logs
├── README.md          # This file
├── CHANGELOG.md       # Version history
├── KNOWN_ISSUES.md    # Known issues and limitations
├── docs/              # Documentation
│   ├── GUIDELINES.md  # Development guidelines
│   ├── PROJECT_PLAN.md # Project roadmap
│   ├── TESTING_GUIDE.md # Testing instructions
│   └── archive/       # Archived documentation
└── RELEASE_NOTES_v1.4.0.md  # Latest release notes

🎯 System Requirements

Minimum:

Python 3.8+
4GB RAM
ComfyUI instance
Discord Bot Token

Performance:

Uptime: >99% reliability
Response Time: <30s images, <10min videos
Concurrent Users: Unlimited
File Management: Automatic cleanup
Error Recovery: Graceful handling

🤝 Contributing

We welcome contributions! Here's how to get started:

Fork the repository
Create feature branch: git checkout -b feature/amazing-feature
Follow guidelines: Check GUIDELINES.md
Test thoroughly: Ensure all features work
Submit PR: With detailed description

Development Guidelines:

Follow modular architecture
Add comprehensive error handling
Include type hints and docstrings
Test with real ComfyUI workflows
Update documentation

See docs/GUIDELINES.md for detailed guidelines.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

ComfyUI Team - For the amazing AI generation framework
Discord.py Developers - For the excellent Discord library
AI Art Community - For inspiration and feedback
Contributors - Everyone who helps improve DisComfy

🆘 Support

Need help? Here's how to get support:

📖 Check Documentation - Most issues are covered here
🔍 Search Issues - Someone may have had the same problem
🐛 Report Bugs - Create detailed issue reports
💬 Join Community - Get help from other users
📧 Contact Maintainers - For complex issues

🎨 Ready to create amazing AI art? Install DisComfy and let your creativity flow! ✨

DisComfy v2.2.0 - The Professional Discord ComfyUI Bot

📚 Documentation

README.md - This file (getting started, features, usage)
CHANGELOG.md - Complete version history
KNOWN_ISSUES.md - Known issues and limitations
RELEASE_NOTES_v1.4.0.md - Latest release notes
docs/GUIDELINES.md - Development guidelines
docs/PROJECT_PLAN.md - Project roadmap
docs/TESTING_GUIDE.md - Testing instructions

For archived documentation and research, see docs/archive/.

Name		Name	Last commit message	Last commit date
Latest commit History 91 Commits
.github/workflows		.github/workflows
bot		bot
config		config
core		core
docs		docs
tests		tests
utils		utils
wiki-temp		wiki-temp
workflows		workflows
.dockerignore		.dockerignore
.gitignore		.gitignore
CHANGELOG.md		CHANGELOG.md
CLAUDE.md		CLAUDE.md
Dockerfile		Dockerfile
KNOWN_ISSUES.md		KNOWN_ISSUES.md
LICENSE		LICENSE
README.md		README.md
RELEASE_NOTES.md		RELEASE_NOTES.md
config.example.json		config.example.json
config.py		config.py
env.example		env.example
main.py		main.py
push-docker.sh		push-docker.sh
requirements.txt		requirements.txt
setup.py		setup.py

Folders and files

Latest commit

History

Repository files navigation

🎨 DisComfy - Discord AI Art Bot

🎉 What's New - Qwen Image 2512 Model

📝 Previous Releases

🌟 Why Choose DisComfy?

🔥 Key Highlights:

✨ Core Features

🎨 AI Image Generation

🔍 Advanced Image Upscaling

🎬 Professional Video Generation

✏️ Advanced Image Editing

📊 Real-Time Progress Tracking

⚙️ Interactive User Experience

🛡️ Production-Ready Features

🚀 Real-Time Progress Tracking

📈 Progress Display:

🔧 Technical Features:

⚙️ Interactive Parameter Selection

🔍 Upscale Customization:

🎬 Video Customization:

🎯 Smart Defaults:

🚀 Quick Start Guide

Prerequisites:

1. Clone & Setup:

2. Configure Bot:

3. Discord Bot Setup:

4. Prepare ComfyUI:

5. Launch Bot:

🐳 Docker Quick Start

Using the Pre-built Image:

Running with Docker:

Docker Compose Example:

Building from Source:

📖 Usage Examples

Basic Image Generation:

Image Editing (Flux Kontext - High Quality):

Image Editing (Qwen - Ultra Fast):

Multi-Image Qwen Editing (NEW in v1.3.0):

Advanced Parameters:

Using LoRAs:

Post-Generation Actions:

🔧 Advanced Configuration

Custom Workflows:

Enable/Disable Workflows:

Performance Tuning:

🛠️ Troubleshooting

Common Issues:

Getting Help:

📁 Project Structure

🎯 System Requirements

Minimum:

Recommended:

Performance:

🤝 Contributing

Development Guidelines:

📄 License

🙏 Acknowledgments

🆘 Support

📚 Documentation

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases 21

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

Packages