Rate limits

Rate limits apply per user. All of your API keys share a single budget, and so do the Solana regions — eu. and us. draw on one allowance rather than one each.

Window	Limit
Per minute	60 requests
Per hour	1,200 requests
Per week	10,000 requests

Response headers

Every response (including errors) includes your current budget:

Header	Description
`X-RateLimit-Limit-Minute`	Configured per-minute limit
`X-RateLimit-Limit-Hour`	Configured per-hour limit
`X-RateLimit-Limit-Week`	Configured per-week limit
`X-RateLimit-Remaining-Minute`	Requests remaining in the current minute window
`X-RateLimit-Remaining-Hour`	Requests remaining in the current hour window
`X-RateLimit-Remaining-Week`	Requests remaining in the current week window

Counts are shared across regions and reconciled continuously, so the Remaining-* values are accurate within a moment rather than to the request. Right after you move traffic to another region its headers can read high until they catch up. A sudden increase in remaining budget is stale reporting, not new headroom — if you pace against these headers, keep your own counter as a floor.

HTTP 429: Too Many Requests

When any window is exhausted the server responds with 429 (code RATE_LIMITED) and adds a Retry-After header:

Header	Description
`Retry-After`	Seconds until the blocking window resets

The details field contains retry_after=<seconds>. The rejected request is not counted against your budget.

Getting Started

Core Concepts

Response headers

HTTP 429: Too Many Requests

​Response headers

​HTTP 429: Too Many Requests

Response headers

HTTP 429: Too Many Requests