OpenAI Integration

The OpenAI integration allows you to leverage OpenAI’s powerful AI models including GPT-4, DALL-E, Whisper, and Sora for a variety of tasks including text generation, image generation and analysis, audio transcription and generation, and video generation.

Available Nodes

n8n provides two types of OpenAI nodes:

OpenAI Node

Specialized operations: image generation (DALL-E), audio (Whisper, TTS), and video generation (Sora)

OpenAI Chat Model

For text generation, reasoning, and tools - use with AI Agent

For most text generation and LLM tasks, use the AI Agent node with OpenAI Chat Model instead of the standalone OpenAI node. The OpenAI node is optimized for specialized operations like image, audio, and video generation.

Prerequisites

Before you begin, you’ll need:

An OpenAI account
An OpenAI API key (get one from OpenAI Platform)
Sufficient API credits in your OpenAI account

Setup

Create OpenAI API Credentials

Go to the OpenAI API Keys page
Click “Create new secret key”
Give your key a name and copy it immediately (you won’t be able to see it again)
Store the key securely

Add Credentials in n8n

In your n8n workflow, add an OpenAI node
Click on the Credential to connect with dropdown
Click Create New Credential
Paste your API key
(Optional) Configure custom base URL if using a proxy
Click Save

Test the Connection

Test your credentials by running a simple operation to ensure everything is configured correctly.

OpenAI Node

The OpenAI node provides specialized operations for image, audio, and video generation. It supports multiple resources and operations.

Available Resources

Generate text responses and perform classifications using GPT models.Operations:

Message: Send messages to GPT models and get responses
Classify: Classify text using moderation models

Example Use Cases:

Generate text completions
Content moderation
Multi-turn conversations
Structured output with JSON mode

Generate and analyze images using DALL-E and GPT-4 Vision.Operations:

Generate: Create images from text descriptions using DALL-E
Analyze: Analyze images using GPT-4 Vision
Edit: Edit existing images with DALL-E 2

Supported Models:

DALL-E 3 (latest, highest quality)
DALL-E 2 (classic, more variations)
GPT Image 1 (new image generation model)

Example:

{
  "resource": "image",
  "operation": "generate",
  "model": "dall-e-3",
  "prompt": "A serene mountain landscape at sunset",
  "options": {
    "size": "1024x1024",
    "quality": "hd",
    "style": "vivid"
  }
}

Generate speech and transcribe audio using Whisper and TTS.Operations:

Generate: Text-to-speech conversion
Transcribe: Convert speech to text (Whisper)
Translate: Translate audio to English

TTS Voices:

alloy, echo, fable, onyx, nova, shimmer

Whisper Models:

whisper-1 (high accuracy speech recognition)

Example:

{
  "resource": "audio",
  "operation": "transcribe",
  "model": "whisper-1",
  "binaryPropertyName": "data"
}

Advanced Features

Function Calling / Tools

Connect custom n8n tools to the OpenAI node to enable function calling:

Add tool nodes to your workflow (e.g., HTTP Request Tool, Code Tool)
Connect them to the Tools input on the OpenAI node
OpenAI will automatically call these tools when needed

Structured Output

Use JSON mode or response format to get structured data:

{
  "responseFormat": "json_object",
  "prompt": "Extract the name, email, and phone from this text as JSON: ..."
}

When using JSON mode, include the word “json” in your prompt and use models released after November 2023.

Streaming Responses

Enable streaming for real-time responses:

{
  "options": {
    "stream": true
  }
}

OpenAI Chat Model

The OpenAI Chat Model node is designed for use with LangChain components, particularly the AI Agent.

Configuration

Add Chat Model Node

Add the OpenAI Chat Model node to your workflow.

Select Model

Choose from available models:

GPT-4 Turbo: Most capable, best for complex tasks
GPT-4: High capability, balanced performance
GPT-3.5 Turbo: Fast and cost-effective
o1/o3 models: Advanced reasoning capabilities

Configure Options

Set temperature, max tokens, and other parameters:

Temperature: Controls randomness (0-2)
Max Tokens: Limit response length
Top P: Nucleus sampling parameter
Frequency Penalty: Reduce repetition
Presence Penalty: Encourage new topics

Connect to AI Agent

Connect the Chat Model to an AI Agent node:

Model Selection

Model	Best For	Context Window	Capabilities
GPT-4 Turbo	Complex reasoning, latest features	128K tokens	Vision, JSON mode, function calling
GPT-4	Balanced performance	8K tokens	High accuracy, reliable
GPT-3.5 Turbo	Speed and cost	16K tokens	Fast responses, good for simple tasks
o1-preview	Advanced reasoning	128K tokens	Complex problem solving
o3-mini	Efficient reasoning	128K tokens	Cost-effective reasoning

Response Formats

{
  "responseFormat": "text"
}

Common Use Cases

1. Content Generation

Generate blog posts, product descriptions, or marketing copy:

2. Image Analysis Pipeline

Analyze images and take actions based on content:

3. Audio Transcription

Transcribe audio files and process the text:

4. AI Agent with Tools

Create an intelligent agent that can use multiple tools:

Best Practices

Choose the Right Model

Use GPT-4 for complex reasoning and high-quality outputs
Use GPT-3.5 Turbo for speed and cost efficiency
Use specialized models (DALL-E, Whisper) for specific tasks

Optimize Token Usage

Set appropriate max_tokens limits
Use system messages to set context efficiently
Consider caching responses for repeated queries

Handle Errors Gracefully

Implement retry logic for rate limits
Check error responses and handle them appropriately
Monitor API usage and costs

Use Streaming for Long Responses

Enable streaming for better user experience
Process chunks as they arrive
Handle connection interruptions

Secure Your API Keys

Never commit API keys to version control
Use n8n’s credential system
Rotate keys regularly
Monitor usage for anomalies

Troubleshooting

Rate Limits

If you encounter rate limit errors:

Implement exponential backoff
Upgrade your OpenAI plan
Use batch processing where possible
Cache responses to reduce API calls

Token Limits

If responses are cut off:

Increase max_tokens parameter
Split large inputs into chunks
Use models with larger context windows
Summarize previous context

Model Not Found

If a model is unavailable:

Check your OpenAI account access level
Verify the model name is correct
Ensure your API key has access to the model
Check OpenAI’s status page for outages

Integration Catalog

AI Integrations

OpenAI Integration

OpenAI Integration

Available Nodes

OpenAI Node

OpenAI Chat Model

Prerequisites

Setup

OpenAI Node

Available Resources

Advanced Features

Function Calling / Tools

Structured Output

Streaming Responses

OpenAI Chat Model

Configuration

Model Selection

Response Formats

Common Use Cases

1. Content Generation

2. Image Analysis Pipeline

3. Audio Transcription

4. AI Agent with Tools

Best Practices

Troubleshooting

Rate Limits

Token Limits

Model Not Found

Resources

Integration Catalog

AI Integrations

​OpenAI Integration

​Available Nodes

OpenAI Node

OpenAI Chat Model

​Prerequisites

​Setup

​OpenAI Node

​Available Resources

​Advanced Features

​Function Calling / Tools

​Structured Output

​Streaming Responses

​OpenAI Chat Model

​Configuration

​Model Selection

​Response Formats

​Common Use Cases

​1. Content Generation

​2. Image Analysis Pipeline

​3. Audio Transcription

​4. AI Agent with Tools

​Best Practices

​Troubleshooting

​Rate Limits

​Token Limits

​Model Not Found

​Resources

OpenAI Integration

Available Nodes

Prerequisites

Setup

OpenAI Node

Available Resources

Advanced Features

Function Calling / Tools

Structured Output

Streaming Responses

OpenAI Chat Model

Configuration

Model Selection

Response Formats

Common Use Cases

1. Content Generation

2. Image Analysis Pipeline

3. Audio Transcription

4. AI Agent with Tools

Best Practices

Troubleshooting

Rate Limits

Token Limits

Model Not Found

Resources