Leon's Utility and API Integration Nodes

ComfyUI Leon API Nodes

A comprehensive collection of ComfyUI custom nodes for AI image generation, LLM chat completions, Midjourney proxy integration, and utility functions via various APIs including HyprLab.

Overview

This package provides multiple categories of nodes to enhance your ComfyUI workflows:

• 🎨 Image Generation APIs: Google Imagen, Luma AI, FLUX, FLUX Kontext
• 🤖 LLM APIs: Chat completions, JSON extraction, model management
• 🎭 Midjourney Proxy: Generate, describe, and upload images
• 🔧 Utility Nodes: Image processing, string manipulation, file uploads

Installation

1. Clone or copy this repository to your ComfyUI custom nodes directory:

   cd ComfyUI/custom_nodes/
   git clone <repository-url> comfyui-leon-nodes
   

2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Restart ComfyUI

🎨 Image Generation Nodes

Supported Image Models

• Google

  • nano-banana
  • imagen-4-ultra, imagen-4, imagen-4-fast
  • imagen-3, imagen-3-fast

• Luma AI

  • photon, photon-flash

• FLUX

  • FLUX 1.1 Pro Ultra, FLUX 1.1 Pro
  • FLUX Pro Canny
  • FLUX Dev, FLUX Schnell
  • FLUX Krea Dev

• FLUX Kontext

  • flux-kontext-max, flux-kontext-pro, flux-kontext-dev

• ByteDance

  • seedream-4
  • dreamina-3.1
  • seedream-3
  • seededit-3

• Ideogram

  • ideogram-v2, ideogram-v2-turbo

• Recraft

  • recraft-v3

• Wan

  • wan-2.7-image
  • wan-2.7-image-pro

• Qwen

  • qwen-image-2
  • qwen-image-2-pro

🤖 LLM API Nodes

Leon LLM Chat API 🤖

Text and vision chat completions with various LLM models.

Features:

• Support for GPT-4o, Claude, Gemini, and more
• Vision capabilities (text + image input)
• Configurable temperature, top_p, max_tokens
• System message support

Usage Example:

Model Selector → LLM Chat API → Text Output
     ↑              ↑
  "gpt-4o"    "Analyze this image"
                     ↑
                Image Input

Leon LLM JSON API 🤖

Structured data extraction with JSON schema validation.

Features:

• JSON schema-based output
• Data extraction and structuring
• Lower temperature for consistency
• Schema validation

Usage Example:

Schema: {
  "type": "object",
  "properties": {
    "email": {"type": "string"},
    "name": {"type": "string"}
  }
}

Leon Model Selector 🔧

Dynamic model selection with live configuration.

Features:

• Dropdown with popular models
• Custom model input override
• Model information display
• Context length awareness

Leon Model Config Loader ⚙️

Fetch and manage available models from API.

Features:

• Live model list fetching
• Cached fallback configuration
• Save to file option
• Error handling with graceful fallback

🎭 Midjourney Proxy Nodes

Leon Midjourney API Generate 🤖

Generate images using Midjourney via proxy server.

Features:

• MID_JOURNEY and NIJI_JOURNEY bot types
• Base64 image array support
• Account filtering
• Polling with timeout handling

Setup:

1. Run Midjourney proxy server (e.g., localhost:8080)
2. Configure API key and endpoint
3. Use prompts with Midjourney syntax (--v 7, --ar 1:1)

Leon Midjourney API Describe 🤖

Get text descriptions of images using Midjourney.

Features:

• Upload image and get 4 description variants
• Automatic description parsing
• Same proxy server integration

Leon Midjourney API Upload 🤖

Upload images to Discord via Midjourney proxy.

Features:

• Convert ComfyUI images to Discord URLs
• Account filtering support
• Direct integration with Midjourney workflows

🔧 Utility Nodes

Leon Image Split 4-Grid 🤖

Split images into 4 quadrants (useful for Midjourney grid results).

Outputs: Top-Left, Top-Right, Bottom-Left, Bottom-Right images

Leon String Combine 🤖

Combine two strings with a linking element.

Usage: Useful for building complex prompts or file paths

Leon ImgBB Upload 🤖

Upload images to ImgBB hosting service.

Features:

• Expiration time control
• Direct URL output
• Error handling

Leon Hypr Upload 🤖

Upload images/videos to HyprLab hosting.

Features:

• Multiple input methods (tensor, file path, URL, base64)
• Output format conversion
• Multipart and JSON upload support

🔗 Common Workflow Patterns

Basic Image Generation

Model Selector → Google/Luma/FLUX API → PreviewImage

Vision Analysis

LoadImage → LLM Chat API → Text Output
              ↑
         Model Selector

Midjourney Grid Processing

Midjourney Generate → Image Split 4-Grid → 4x PreviewImage

Data Extraction

Text Input → LLM JSON API → JSON Output
              ↑
         JSON Schema

Image Upload Chain

Generated Image → Hypr Upload → URL → Use in other APIs

🛠️ Advanced Usage

Model Management

Model Config Loader → Save models.json
Model Selector → Use cached or live models

Multi-Modal Workflows

Text + Image → LLM Chat → Analysis
     ↓
Generated Image → Midjourney Describe → Descriptions
     ↓
Upload → Share URLs

Batch Processing

Multiple Prompts → Loop through FLUX API → Collect Results

🔧 Error Handling

All nodes include robust error handling:

• Retry Logic: Exponential backoff (up to 5 attempts)
• Input Validation: Prompt lengths, required fields
• API Response Validation: Proper error messages
• Network Error Handling: Graceful degradation
• Detailed Logging: Console output for debugging

---

Acknowledge: ComfyUI_BillBum_Nodes | Integrated with: HyprLab API