Google Docs

• 1.
  Implement exponential backoff

  When you receive a 429 error, wait before retrying with increasing delays.

• 2.
  Cache responses when possible

  Store frequently accessed data locally to reduce API calls.

• 3.
  Use batch operations

  Combine multiple operations into single requests where supported.

• 4.
  Monitor your usage

  Use the Primrose dashboard to track your API usage and set alerts.

{
  "error": {
    "code": "rate_limit_exceeded",
    "message": "Too many requests",
    "retry_after": 60
  }
}

When rate limited, the response includes a retry_after value indicating how many seconds to wait before retrying.