ComfyUI-gpt5_image_text

ComfyUI OpenAI GPT-5 & Image Analysis Node

<img width="400" height="297" alt="image" src="https://github.com/user-attachments/assets/93581a1c-f9d8-4fdc-b9ca-61a2bd8dd8cc" />

A powerful custom node for ComfyUI that integrates OpenAI's latest multimodal models. This node allows you to use GPT-5, GPT-4.1, and o1 series models to analyze images, generate captions, or perform complex reasoning tasks directly within your ComfyUI workflows.

✨ Features

• Latest Model Support: Includes support for the GPT-5 family (5.2, Pro, Mini, Nano) and GPT-4.1.
• Reasoning Model Logic: Automatically handles the differences between standard models and "Reasoning" models (GPT-5/o1):
  • Switches between max_tokens and max_completion_tokens automatically.
  • Disables temperature for reasoning models to prevent API errors.
• Multimodal Input: Accepts images and text prompts simultaneously.
• Cost Efficiency: Includes "Mini" and "Nano" model variants for faster, cheaper inference.

🚀 Supported Models

GPT-5 Series (Reasoning & Agents)

• GPT-5.2: Best for coding and agentic tasks.
• GPT-5.2 Pro: High-precision reasoning.
• GPT-5 Mini: Cost-efficient reasoning.
• GPT-5 Nano: Fastest, most lightweight model.

GPT-4 Series

• GPT-4.1: Latest standard flagship.
• GPT-4o: Omni model (great for general image understanding).
• GPT-4o Mini: Fast and cheap standard model.

o1 Series

• o1-preview / o1-mini: Previous generation reasoning models.

📦 Installation

1. Navigate to your ComfyUI custom nodes directory:

   cd ComfyUI/custom_nodes/
   

2. Clone this repository:

   git clone [https://github.com/your-username/ComfyUI-OpenAI-GPT5.git](https://github.com/your-username/ComfyUI-OpenAI-GPT5.git)
   

3. Install the required dependencies:

   cd ComfyUI-OpenAI-GPT5
   pip install -r requirements.txt
   

4. Restart ComfyUI.

⚙️ Usage

1. Add the Node: Right-click -> openai -> analysis -> GPT5 Image & Text.
2. API Key: Paste your OpenAI API Key into the openai_key widget.
3. Connect Image (Optional): Connect a generic image output to the image input slot.
4. Select Model: Choose your desired model from the dropdown.

Parameter Guide

| Parameter | Description | | :--- | :--- | | model | Select the specific OpenAI model (e.g., gpt-5.2, gpt-4o). | | prompt | The main instruction or question for the AI (e.g., "Describe this image"). | | system_prompt | Sets the behavior of the assistant (e.g., "You are a creative writer"). | | temperature | Controls randomness. Note: Ignored for GPT-5 and o1 models (defaults to 1). | | max_tokens | Output limit for standard models (GPT-4o, GPT-4.1). | | max_completion_tokens | Output limit for Reasoning models (GPT-5, o1). |

⚠️ Important Notes

• Reasoning Models (GPT-5 / o1): These models do not support the temperature parameter. The node automatically ignores your temperature setting when these models are selected to prevent API errors.
• Token Limits: If you are using a Reasoning model and the output gets cut off or fails silently, try increasing max_completion_tokens.

📄 License

MIT License