API Integration Overview

What is API Integration?

The API Call node allows you to integrate your conversational flows with any external REST API. Use it to fetch data, send information, or trigger actions in third-party services.

With API integration, your flows can:

• Fetch user data from your CRM or database
• Create records in external systems
• Send notifications through third-party services
• Integrate with AI services like OpenAI
• Process payments through payment gateways
• Trigger workflows in other applications

Key Features

Quick Start

Here's a simple example of fetching user data from an API:

Step-by-Step Setup

1. Create an API Endpoint: Open the Resources menu in the Flow Editor toolbar and select "API Endpoints" to configure your endpoint with base URL and authentication
2. Add an API Call Node: Drag the API Call node into your flow and select your configured endpoint
3. Configure the Request: Set the HTTP method (GET, POST, etc.), path, headers, and request body if needed
4. Store the Response: Choose a variable name to store the API response (e.g., user_data)
5. Use the Response: Access response data in subsequent nodes using Handlebars syntax

Variables in the Path and Body

A variable in the path is URL-encoded, so its value stays within that one part of the path. A value of . or .. stops the request.

Triple braces, such as {{{api.body.next_page}}}, insert the value as-is, without encoding. Use them only for trusted values your flow controls, such as a pagination link returned by the API, and never for anything a visitor typed. The request still has to stay on the endpoint's host.

In a JSON body, a variable inside quotes, such as {"name": "{{customer_name}}"}, is sent as text with any quotes in it escaped. A variable outside quotes, such as {"quantity": {{quantity}}}, stands for a whole number, true/false, list or object, and the request is not sent if it holds anything else.

Working with Response Variables

The response from an API call is stored in a variable that you specify. You can access this data in subsequent nodes using Handlebars templates:

Name: {{user_data.body.name}}
Email: {{user_data.body.email}}
Status: {{user_data.status}}

Response Variable Structure

Each API response includes the following properties:

{
  "status": "success" | "error" | "failure" | "exception",
  "status_code": 200,
  "body": { ... },  // Parsed JSON or plain text
  "headers": { ... },
  "timestamp": "2026-01-06T10:30:00Z",
  "raw": "...",  // Only if includeRaw is enabled
  "error": "..."  // Only present on errors
}

• status: Indicates the result (success, error, failure, or exception)
• status_code: HTTP status code from the API
• body: Parsed response body (automatically parsed for JSON responses)
• headers: Response headers from the API
• timestamp: When the request was made
• raw: Original response text (only if "Include Raw Response" is enabled)
• error: Error message (only present when an error occurs)

Response Caching

The API Call node can cache successful responses within a conversation, avoiding redundant external requests when the same parameters are used more than once in a flow.

How It Works

When caching is enabled, the node computes a fingerprint of each request (HTTP method + URL + query params + body). If a previous successful response with the same fingerprint exists and is within the TTL, it is returned immediately without making a network request.

Enabling Caching

In the API Call node's property panel, open the Response Caching section:

1. Toggle Enable Response Caching on
2. Set Cache Duration: how many seconds a cached response remains valid (default: 300 = 5 minutes; range: 60–86400)

Important Constraints

Related Resources