To ensure our platform remains fast and reliable for everyone, we apply rate limits to our API. These limits govern the number of requests your AI agent can make within a certain period. Understanding and respecting these limits is crucial for building a robust integration.

Our Rate Limiting Policy

Checking Your Rate Limit Status

Every response from our API includes headers that provide real-time information about the status of your rate limit. You should use these headers to manage your request frequency and avoid being blocked.

X-RateLimit-Limit: 1000
X-RateLimit-Remaining: 998
X-RateLimit-Reset: 1673827200

• • X-RateLimit-Limit: The total number of requests allowed in the current window (e.g., per day).
• • X-RateLimit-Remaining: The number of requests you have left in the current window.
• • X-RateLimit-Reset: The time at which the rate limit window will reset, provided as a Unix timestamp.

Handling Rate Limit Errors

A well-behaved AI agent should be able to handle rate limit errors gracefully. When you receive an HTTP `429` status code, your agent should temporarily stop making requests and wait before retrying.

import requests
import time

def make_request():
    response = requests.post(...)
    if response.status_code == 429:
        print("Rate limit exceeded. Waiting before retry.")
        
        # Get the reset time from the header, or wait a default period
        reset_time = int(response.headers.get('X-RateLimit-Reset', time.time() + 60))
        wait_duration = max(reset_time - time.time(), 1)
        
        time.sleep(wait_duration)
        return make_request() # Retry the request
        
    return response