Rate Limiting & Throttling

Blueprint applies a usage plan to Partner API requests to keep the API reliable and available.

Default usage plan

The rate and burst limits work together. The burst bucket can hold up to 120 requests and refills
at 2 requests per second. Each request consumes capacity from the bucket. This allows a short burst
of traffic above 2 requests per second while capacity is available. After the bucket is depleted,
requests can continue at the sustained rate of 2 requests per second as capacity refills.

The daily quota applies independently of the burst bucket. Requests may be throttled after the
10,000-request daily quota is reached, even when burst capacity is available.

Handling throttled requests

When a request exceeds an applicable limit, the API returns HTTP 429 Too Many Requests. Clients
should:

• Pace requests instead of sending large batches concurrently
• Retry 429 responses using exponential backoff with jitter
• Prefer webhooks over frequent polling for clinical artifacts
• Avoid automatically retrying create requests without first checking whether the resource was
  created, because the Partner API does not support idempotency keys

For more information about safe retries, see
Pagination & Limits.

Requesting higher limits

If you expect your integration to require a higher rate, burst limit, or daily quota, contact
Blueprint to request an increase to your usage plan.