Rate limits

Request limits and error handling

To ensure stable performance and prevent misuse, our API enforces a rate limit per tenant, regardless of how many API keys a tenant uses.

How it works

• Each tenant is allowed up to 20 API requests per minute, across all their API keys combined.

• We use a sliding window strategy:

  • The system continuously tracks the number of requests made by a tenant over the last 60 seconds.

  • Requests are allowed as long as the tenant has not exceeded the 20-request limit within that moving window.

  • As soon as an older request drops out of the window, a new request is allowed.

Limit details

Setting	Value
Scope	Per tenant
Shared across	All API keys per tenant
Limit	20 request
Window	1 minute (sliding)

When limit is exceeded

If a tenant exceeds the limit:

• The API responds with HTTP 429 – Too Many Requests.

• A Retry-After header is included in the response, indicating how long to wait before retrying.

Best practices

• Distribute requests evenly rather than sending many in bursts.

• If you receive a 429 error, pause requests based on the Retry-After header before retrying.

• Design your integration with automatic retry logic (e.g. exponential backoff) to handle throttling gracefully.