Alloy uses API rate limits to manage request volume and help maintain service quality. Limits are generally defined by endpoint and time window, rather than as one universal limit for every API request.
Customer-specific configurations
Standard rate limits provide a general reference, but your organization’s configuration may be different. Limits can vary by endpoint, HTTP method, environment, or account-specific requirements. Always use the rate-limit headers returned by the API for specific information about your organization.
Monitor your rate limits
Alloy returns rate-limit information in API response headers, including:
alloy-ratelimit-limit: The total number of requests allowed in the current window.
alloy-ratelimit-remaining: The number of requests remaining in the current window.
alloy-ratelimit-reset: When the current rate-limit window is expected to reset.
retry-after: How long to wait before retrying, when provided.
If you exceed a limit
When your request volume exceeds the configured limit, Alloy returns a 429 response. Requests are rejected until the rate-limit window resets.
For more information, see Request Handling and Retries.
Request help with rate limits
If you regularly receive 429 responses or need to support a temporary increase in traffic, contact Alloy Support. Include the affected endpoint and HTTP method, environment, approximate request volume and timing, relevant response headers, and the error response. Rate-limit changes are reviewed based on the use case and are not guaranteed.