Skip to content
CentralMind

ChatGPT Integration

This guide explains how to integrate your Gateway-generated API with ChatGPT by creating a custom GPT with actions.

Prerequisites

Before integrating with ChatGPT, ensure you have:

  1. A running Gateway API with a publicly accessible endpoint
  2. OpenAPI/Swagger documentation available at your API endpoint (e.g., https://your-api.com/swagger/)
  3. A ChatGPT Plus subscription to access the GPT Builder
  4. Necessary authentication information if your API requires it

Creating a Custom GPT with Gateway API Actions

Step 1: Access GPT Builder

  1. Log in to your ChatGPT Plus account
  2. Navigate to GPT Builder at https://chatgpt.com/gpts/mine

img

Step 2: Set Up Your Custom GPT

  1. Click “Create new GPT”
  2. Fill in the basic information:
    • Name: Choose a name related to your API’s purpose
    • Description: Explain what your API does and how users can interact with it
    • Instructions: Provide detailed guidance on how the GPT should use your API

img

Step 3: Add Actions to Connect to Your Gateway API

  1. In the GPT Builder interface, navigate to the “Actions” tab

  2. Click “Create new action”

  3. Enter a name for your action that describes what it does

  4. For “Authentication,” select the appropriate method:

    • “None” if your API is publicly accessible
    • “API Key” if your API uses key-based authentication
    • “OAuth” for OAuth authentication

Step 4: Configure OpenAPI Specification

  1. Select “OpenAPI schema” for Schema type

  2. Enter your Gateway API’s OpenAPI URL (e.g., https://your-gateway-api.com/openapi.json or https://your-gateway-api.com/swagger/)

    Alternatively, you can:

    • Click “Import from URL” and enter your OpenAPI URL
    • Copy and paste the OpenAPI JSON directly into the schema field

img

Step 5: Customize Privacy and Settings

  1. Review the endpoints that will be accessible to the GPT
  2. Configure which operations your GPT can perform
  3. If required, set up authentication details:
    • For API Key authentication, specify the header name (e.g., X-API-KEY)
    • Enter placeholder values that users will need to provide

Step 6: Save and Test Your Custom GPT

  1. Click “Save” to create your action
  2. Test your API connection with example queries
  3. Make adjustments to the instructions as needed to optimize the interaction

Example Conversation Flow

Here’s an example of how a conversation with your Gateway API-enabled GPT might look:

User: “Can you show me pricing for instance c6g.2xlarge and specification?” img

Troubleshooting

If you encounter issues with your ChatGPT integration:

  1. API not accessible: Ensure your API is publicly accessible from the internet
  2. Authentication errors: Verify your authentication details are correctly configured
  3. Schema errors: Check that your OpenAPI specification is valid and properly formatted
  4. Rate limiting: Consider implementing rate limiting on your API to prevent abuse

Publishing Your Custom GPT (Optional)

If you want to share your GPT with others:

  1. Go to the “Configure” tab and select visibility settings
  2. Add appropriate usage instructions and examples
  3. Click “Publish” to make your GPT available according to your selected visibility

[IMAGE PLACEHOLDER: Screenshot of the publishing interface with visibility options]

For more details on creating custom GPTs, refer to the OpenAI GPT Builder Documentation.