ComfyUI-Freepik

ComfyUI-Freepik

Professional Freepik AI integration for ComfyUI workflows

Bring Freepik's powerful AI capabilities directly into your ComfyUI workflows with seamless integration, smart caching, and pay-as-you-go pricing.

✨ Features

• 🎨 Mystic AI - Photorealistic text-to-image generation
• 🔍 Creative Upscaler - Magnific.ai powered prompt-guided enhancement (2x-16x)
• 🎯 Precision Upscaler - High-fidelity upscaling without hallucinations
• ✂️ Background Removal - Instant AI-powered background removal
• 💾 Smart Caching - Automatic result caching (30-day retention)
• 💰 Cost Management - Real-time cost estimation before execution
• 🔄 Async Processing - Non-blocking task management with progress tracking

🚀 Quick Start

Installation

1. Clone or download into your ComfyUI custom nodes directory:

cd ComfyUI/custom_nodes
git clone https://github.com/yourusername/ComfyUI-Freepik.git
cd ComfyUI-Freepik

2. Install dependencies:

pip install -r requirements.txt

3. Get your API key:

   • Visit Freepik API Dashboard
   • Sign up and create an API key
   • Get $5 free credits to start

4. Restart ComfyUI and look for nodes under Freepik/ category

First Generation

1. Add Freepik Mystic node
2. Enter your API key
3. Write a prompt
4. Connect to Preview Image or Save Image
5. Queue prompt!

📦 Available Nodes

🎨 Freepik Mystic (Text-to-Image)

Generate photorealistic 2K images from text prompts.

Features:

• Multiple aspect ratios (16:9, 1:1, 9:16, 4:3, 21:9)
• LoRA style support
• Seed control for reproducibility
• Cost: €0.10 per image

Outputs:

• square_1_1: 2048×2048
• widescreen_16_9: 2752×1536
• social_story_9_16: 1536×2752

🔍 Freepik Upscaler Creative

AI-powered upscaling with prompt guidance (Powered by Magnific.ai)

Features:

• Upscale factors: 2x, 4x, 8x, 16x
• Prompt-guided enhancement
• Creativity control (0.0-1.0)
• Detail level control (0.0-1.0)
• Cost: €0.10 - €1.60+ depending on output size

Best for:

• Architectural renders (creativity: 0.3-0.5)
• Interior design (detail: 0.6-0.7)
• Concept art enhancement

🎯 Freepik Upscaler Precision

High-fidelity upscaling without AI hallucinations

Features:

• Faithful to original content
• Perfect for logos, UI, technical drawings
• Denoise strength control
• Sharpen control
• Same cost structure as Creative

✂️ Freepik Remove Background

Instant background removal with alpha mask output

Features:

• Returns image + alpha mask
• Fast and reliable
• Minimal cost

💰 Pricing & Cost Comparison

Pay-As-You-Go (No Subscription Required)

| Operation | Cost | |-----------|------| | Text-to-Image (2K) | €0.10 | | Upscale 2x | €0.10 | | Upscale 4x | €0.40 | | Upscale 8x | €1.60 | | Remove Background | Low/Free |

vs Magnific AI Platform

ComfyUI-Freepik is more cost-effective when:

• ✅ You process < 90 upscales/month → Save 33-89%
• ✅ You need automation and complex workflows
• ✅ Usage is variable month-to-month
• ✅ You prefer pay-as-you-go with no monthly commitment

Breakeven Analysis:

| Monthly Volume | Magnific Subscription | API Pay-as-you-go | Savings | |----------------|----------------------|-------------------|---------| | 10 upscales | $39 | $4.32 | $34.68 (89%) | | 50 upscales | $39 | $21.60 | $17.40 (45%) | | 90 upscales | $39 | $38.88 | Breakeven | | 200 upscales | $39 | $86.40 | -$47.40 |

For most users (< 90/month): ComfyUI-Freepik saves 33-89%

📊 Read full cost analysis

🎯 Example Workflows

Basic Generation

Freepik Mystic → Preview Image

Generation + Enhancement

Freepik Mystic (2K, 16:9) → 
Freepik Upscaler Creative (4x) → 
Save Image

Output: 2752×1536 → 11008×6144

Architectural Visualization Pipeline

Freepik Mystic (architectural prompt) → 
Freepik Upscaler Precision (2x-4x) → 
Post-processing nodes → 
Final delivery

Background Removal

Load Image → 
Freepik Remove Background → 
Composite with new background

🏗️ Architecture

ComfyUI_Freepik/
├── __init__.py              # Node registration
├── requirements.txt         # Dependencies
├── api/
│   ├── client.py           # HTTP client with retry logic
│   └── tasks.py            # Async task manager
├── nodes/
│   ├── generation/
│   │   └── mystic_node.py  # Text-to-image
│   ├── editing/
│   │   └── upscaler_nodes.py # Upscalers
│   └── utilities/
│       └── remove_bg_node.py # Background removal
└── utils/
    ├── cache.py            # Caching system
    └── image_utils.py      # Image conversions

🔧 Technical Highlights

• Smart Caching: MD5-based parameter matching with automatic cleanup
• Async Processing: Non-blocking execution with configurable intervals
• Error Recovery: Exponential backoff retry strategy (3 retries)
• Multi-format Support: JSON + multipart/form-data requests
• Progress Tracking: Real-time status updates during processing
• Cost Transparency: Estimation before execution

📋 Requirements

• Python 3.9+
• ComfyUI (latest version recommended)
• Internet connection
• Freepik API key

Python Dependencies:

• requests
• Pillow
• torch

🎨 Parameter Reference

Magnific Creative Upscaler Equivalence

| Freepik Platform | ComfyUI Node | Value Mapping | |-----------------|--------------|---------------| | Creativity (0-10) | creativity | 0.0-1.0 (divide by 10) | | HDR (0-10) | detail_level | 0.0-1.0 (divide by 10) | | Scale Factor | upscale_factor | "2x", "4x", "8x", "16x" | | Prompt | prompt | Direct equivalent |

Recommended for Architecture:

• Creativity: 0.3-0.5
• Detail Level: 0.6-0.7
• Prompt: "enhance architectural details, photorealistic render, natural lighting"

📖 Full parameter reference

🛠️ Advanced Usage

Batch Processing

# Process multiple images automatically
for image in image_list:
    mystic_node.generate(
        api_key=api_key,
        prompt=prompt,
        aspect_ratio="widescreen_16_9"
    )

Custom Workflows

Integrate with other ComfyUI nodes for complex pipelines:

• Pre-processing with ControlNet
• Post-processing with color correction
• Automated client delivery systems

Cost Optimization

• Enable caching to avoid redundant API calls
• Batch similar operations
• Use appropriate resolution for your needs

⚠️ Limitations

API Restrictions

• Mystic outputs fixed ~2K resolution (aspect ratio determines dimensions)
• Some Magnific parameters not available (Resemblance, Fractality, Engine)
• Rate limits apply (check Freepik API documentation)

vs Magnific.ai Platform

• ❌ No Resemblance slider
• ❌ No Fractality slider
• ❌ No Engine selection
• ✅ But: Full workflow automation
• ✅ And: Pay-as-you-go pricing

🗺️ Roadmap

Phase 2 (Future)

• [ ] Video Generation nodes (Kling v2.5, PixVerse V5)
• [ ] Image Relighting
• [ ] Style Transfer
• [ ] Image-to-Prompt utility
• [ ] LoRA Training nodes
• [ ] Batch processing manager
• [ ] Credit tracker dashboard

Improvements

• [ ] Remove debug prints
• [ ] Add architectural preset system
• [ ] Optimize cache management
• [ ] Progress bars in UI
• [ ] Better error messages

📚 Documentation

• Installation Guide
• Cost Analysis
• Parameter Reference
• Changelog

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

MIT License - see LICENSE file for details

🙏 Credits

Developer: Felipe @ Pixelflakes
Built for: Architectural visualization workflows
Powered by: Freepik API + Magnific.ai technology

Special thanks to:

• ComfyUI community
• Freepik API team
• Magnific.ai for upscaling technology

🔗 Links

• Freepik API Documentation
• Freepik Developer Dashboard
• ComfyUI
• Magnific.ai

📞 Support

• Issues: GitHub Issues
• API Docs: https://docs.freepik.com
• Freepik Support: https://support.freepik.com

---

Made with ❤️ for the ComfyUI community

If you find this useful, please ⭐ star the repository!