Skip to main content
To ensure the stability and fair use of our platform for all users, the Gloo AI API enforces rate limits on incoming requests. These limits define the number of API calls that can be made within a certain period.

​
How Rate Limiting Works

Rate limiting is applied on a per-API-key basis. The specific limits depend on your organization’s subscription plan. When you exceed the number of allowed requests in a given time window, the API will respond with an HTTP 429 Too Many Requests error.

​
Handling Rate Limits Gracefully

When you receive a 429 Too Many Requests error, you should pause your requests until the time specified in the X-RateLimit-Reset header. A common strategy for handling rate limits is to implement an exponential backoff mechanism. This involves waiting for a progressively longer period between retries.

​
Example Error Response

{
  "error": "Rate limit exceeded",
  "code": "too_many_requests"
}

​
Example Exponential Backoff (Pseudocode)

retries = 0
max_retries = 5

loop:
  response = make_api_request()

  if response.status_code == 200:
    // Success!
    break

  if response.status_code == 429:
    if retries < max_retries:
      wait_time = (2 ** retries) + random_jitter()
      sleep(wait_time)
      retries += 1
    else:
      // Max retries reached, handle error
      log("Max retries exceeded for rate limit.")
      break
  else:
    // Handle other errors
    break