n8n-nodes-openai-gpt5

This is an n8n community node that provides access to OpenAI's GPT-5 models and advanced features through the Responses API. It supports PDF processing, multiple file inputs, and all the latest GPT-5 capabilities including reasoning effort control, verbosity settings, and reasoning summaries. Now with AI Agent Tool support!

Features

• 🚀 Full GPT-5 Support: Access GPT-5, GPT-5 Mini, GPT-5 Nano models
• 📄 PDF Processing: Upload and process PDF files directly
• 🎯 Reasoning Control: Fine-tune reasoning effort (minimal, low, medium, high)
• 📝 Verbosity Control: Adjust output length (low, medium, high)
• 🧠 Reasoning Summaries: Get insights into the model's thinking process
• 🔗 Multiple File Support: Process multiple PDFs and images in one request
• ⚡ O-Series Models: Support for O3, O3 Pro, O3 Mini, O4 Mini models
• 🤖 AI Agent Tool: Use as a tool with n8n AI Agent nodes

Installation

In n8n

1. Go to Settings > Community Nodes
2. Search for n8n-nodes-openai-gpt5
3. Click Install

Manual Installation

npm install n8n-nodes-openai-gpt5

Then restart your n8n instance.

Setup

1. Get OpenAI API Key:

   • Sign up at OpenAI Platform
   • Generate an API key from your account settings

2. Configure Credentials in n8n:

   • Go to Credentials > New
   • Select OpenAI GPT-5 API
   • Enter your API key
   • (Optional) Add Organization ID if you have one
   • Save the credentials

Usage

Using as an AI Agent Tool

This node can be used as a tool with n8n's AI Agent nodes, allowing AI agents to leverage GPT-5's advanced reasoning capabilities:

1. Enable Community Tools (if needed):

   • Set environment variable: N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true

2. Connect to AI Agent:

   • Add an AI Agent node to your workflow
   • Add this OpenAI GPT-5 node
   • Connect the GPT-5 node to the AI Agent's tool input

3. AI Tool Mode:

   • The node includes a special "AI Tool Mode" operation
   • Simplified interface optimized for AI agent usage
   • Automatically handles file processing if needed

4. Example Use Cases:

   • Document analysis and summarization
   • Complex reasoning tasks
   • PDF content extraction and Q&A
   • Multi-modal analysis (text + images)

Basic PDF Processing

1. Add the OpenAI GPT-5 node to your workflow
2. Select Upload & Process PDF operation
3. Choose input type:
   • Binary Data: Use files from previous nodes
   • File URL: Direct download from URL
   • File Path: Server file path
4. Enter your prompt
5. Configure additional options as needed

Process with File ID

If you've already uploaded a file and have its ID:

1. Select Process with File ID operation
2. Enter the file ID
3. Provide your prompt

Node Options

Models

• GPT-5: Most capable model
• GPT-5 Mini: Smaller, faster variant
• GPT-5 Nano: Smallest, fastest variant
• GPT-4.1: Previous generation model
• O3: Advanced reasoning model
• O3 Pro: Enhanced O3 with better reasoning
• O3 Mini: Smaller O3 variant
• O4 Mini: Latest mini reasoning model

Reasoning Effort (GPT-5 & O-Series)

• Minimal: Fastest responses with minimal reasoning (GPT-5 only)
• Low: Lower reasoning effort for faster responses
• Medium: Balanced reasoning and speed (default)
• High: Maximum reasoning capability

Verbosity (GPT-5 only)

• Low: Most concise output
• Medium: Balanced verbosity
• High: More detailed output

Reasoning Summary

• None: No reasoning summary
• Auto: Automatic summary generation
• Detailed: Detailed reasoning summary

Note: Reasoning summaries are not guaranteed for every request.

Additional Options

• Max Tokens: Control output length (default: 4096)
• Temperature: Control randomness (0-2, default: 0.7)
• Additional Files: Include multiple files in the request
• Purpose: Specify file upload purpose (default: "user_data")

Example Workflows

1. PDF Analysis

[Read Binary File] → [OpenAI GPT-5: Analyze PDF] → [Output]

2. Multiple Document Processing

[HTTP Request: Get PDFs] → [OpenAI GPT-5: Process with Additional Files] → [Save Results]

3. High-Reasoning Analysis

[Input Data] → [OpenAI GPT-5: High Reasoning + Low Verbosity] → [Slack Message]

API Structure

The node uses OpenAI's Responses API (/v1/responses) with the following structure:

GPT-5 Models

{
  "model": "gpt-5",
  "input": [...],
  "reasoning": {
    "effort": "medium",
    "summary": "auto"
  },
  "text": {
    "verbosity": "medium"
  }
}

O-Series Models

{
  "model": "o3",
  "input": [...],
  "reasoning_effort": "medium"
}

Performance Guidelines

Reasoning Effort	Response Time	Use Case
Minimal	~3-4s	Quick answers, simple queries
Low	~3-5s	Standard queries with basic reasoning
Medium	~5-10s	Balanced performance and quality
High	~15-50s	Complex analysis, critical decisions

Limitations

• Not Supported: Temperature control for reasoning models, image generation with O3 Pro
• Reasoning Summaries: Not guaranteed for every request
• Token Limits: Subject to your OpenAI plan limits
• File Size: PDF uploads subject to OpenAI file size limits

Contributing

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

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Development

# Install dependencies
npm install

# Build the node
npm run build

# Run tests
OPENAI_API_KEY=your-key npm test

# Lint code
npm run lint

Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• n8n Community: n8n Community Forum

License

MIT - See LICENSE file for details

Author

Jez - Jezweb

Changelog

v0.4.0 (2025-01-10)

• Added AI Agent Tool support with usableAsTool property
• New "AI Tool Mode" operation for simplified AI agent usage
• Added codex metadata for better tool discovery
• Enhanced documentation for AI agent integration
• Automatic file handling in AI Tool Mode

v0.3.2 (2025-01-10)

• Temperature parameter now excluded for GPT-5 models (not supported by reasoning models)
• Updated temperature description to clarify it's not supported by GPT-5
• Reordered options alphabetically for better maintainability

v0.3.1 (2025-01-10)

• Fixed response extraction for simplified output
• Improved input structure for text-only requests
• Enhanced error handling with detailed messages
• Fixed response parsing for GPT-5 output structure
• Removed preamble option (only for tool-calling scenarios)

v0.3.0 (2025-01-10)

• Added verbosity control for GPT-5 models
• Added reasoning summary support
• Enhanced reasoning effort with all levels
• Initial error handling improvements

v0.2.2 (2025-01-09)

• Initial public release
• PDF processing support
• Multiple file inputs
• Basic reasoning effort control

Acknowledgments

• Built for the n8n workflow automation platform
• Uses OpenAI API for AI capabilities
• Community feedback and contributions