HTTP Request: Call Any API

Ch06 HTTP Request: Call Any API

The HTTP Request node is n8n's most-used node by a wide margin. It lets you call virtually any REST API — from OpenAI to Stripe, from internal microservices to government open data platforms — without writing a single line of networking code. This chapter covers every configuration option in depth, with complete examples for the most common API call patterns.

Node Configuration Reference

Basic Settings

Method: HTTP method — GET, POST, PUT, PATCH, DELETE, HEAD, OPTIONS.

URL: Target API endpoint; supports n8n expressions for dynamic values, e.g., https://api.github.com/repos/{{ $json.owner }}/{{ $json.repo }}/issues.

Authentication: Built-in options include:

• Predefined Credential Type: Select from saved Credentials (recommended)
• Generic Credential Type: Custom Header Auth, Basic Auth, Bearer Token
• None: Public APIs that require no auth

Headers

Most REST APIs need Content-Type and Authorization headers:

{
  "Content-Type": "application/json",
  "Authorization": "Bearer {{ $vars.API_KEY }}",
  "X-Custom-Header": "custom-value"
}

Body

POST/PUT/PATCH requests typically need a body:

• JSON: Most common, pairs with Content-Type: application/json
• Form Data: application/x-www-form-urlencoded
• Multipart Form Data: File uploads
• Raw/Binary: Raw byte streams

Query Parameters

GET parameters go in Query Parameters — n8n handles URL encoding automatically:

page=1&limit=20&status=active&search={{ $json.keyword }}

Response Handling

Response Format:

• Automatic (default): Auto-detects; JSON is parsed, others become text
• JSON: Force JSON parsing
• Text: Return as a plain string
• File: Return as binary data (for file downloads)

Practical Examples

Example 1: Call the OpenAI Chat API

{
  "method": "POST",
  "url": "https://api.openai.com/v1/chat/completions",
  "headers": {
    "Authorization": "Bearer {{ $vars.OPENAI_API_KEY }}",
    "Content-Type": "application/json"
  },
  "body": {
    "model": "gpt-4o",
    "messages": [
      { "role": "system", "content": "You are a helpful assistant." },
      { "role": "user", "content": "{{ $json.userMessage }}" }
    ],
    "temperature": 0.7,
    "max_tokens": 1000
  }
}

After execution, $json.choices[0].message.content contains the AI response.

Example 2: Create a GitHub Issue

{
  "method": "POST",
  "url": "https://api.github.com/repos/{{ $json.owner }}/{{ $json.repo }}/issues",
  "headers": {
    "Authorization": "token {{ $vars.GITHUB_TOKEN }}",
    "Accept": "application/vnd.github.v3+json"
  },
  "body": {
    "title": "{{ $json.issueTitle }}",
    "body": "{{ $json.issueBody }}",
    "labels": ["bug", "auto-created"]
  }
}

Example 3: Handle Paginated APIs

Many APIs paginate their responses. Use a Loop and Code node to auto-paginate:

// Code node: pagination logic
const staticData = $getWorkflowStaticData('global');
const allItems = staticData.collectedItems || [];

const currentPageItems = items.map(i => i.json);
allItems.push(...currentPageItems);

const hasNextPage = items.length === 100; // Assuming 100 per page
const nextPage = (staticData.currentPage || 1) + 1;

if (hasNextPage) {
  staticData.currentPage = nextPage;
  staticData.collectedItems = allItems;
  return items; // Loop continues
} else {
  // All pages collected — reset and return everything
  staticData.currentPage = 1;
  staticData.collectedItems = [];
  return allItems.map(item => ({ json: item }));
}

Advanced Features

Batch Size: Rate-Limit API Calls

In Options → Batch Size, set how many items to process per batch. Combine with Wait Between Batches to pace API calls and avoid rate limits:

Batch Size: 5
Wait Between Batches: 1000  # 1 second between batches

Timeout and Retry

Configure in Options:

• Timeout: Request timeout in milliseconds (default: 300000 = 5 minutes)
• Retry On Fail: Auto-retry on failure — recommended for unstable external APIs

Allow Unauthorized Certs

For internal services with self-signed certificates, enable Options → Allow Unauthorized Certs (not recommended for public internet services).

Accessing Response Headers

Access response headers via $response.headers:

• Rate limit info: $response.headers['x-ratelimit-remaining']
• Next page URL: $response.headers['link']
• Request ID: $response.headers['x-request-id']

Common Error Reference

Debug tip: Enable Full Response in node Settings to include status code ($json.statusCode), headers ($json.headers), and body ($json.body) in the output — invaluable for debugging non-2xx responses.