Getting started

Rate limits

By default you can make 3,600 requests per hour. Every response tells you how many you have left.

Default limit

The limit applies per account, across all of its tokens, and resets every hour. Every call made with your token counts, whether it succeeds or fails.

Rate limits by plan
Plan	Requests / hour	Sports	Average rate
Trial	1,800	All sports	1 every 2 s
Single sport	3,600	1 sport	1 every 1 s
All sports	3,600	All 28 sports	1 every 1 s
Scale	200,000	All 28 sports	1 every 18 ms

Rate limit headers

Every response, successful or not, includes three headers:

Rate limit headers
Header	Example	Description
X-RateLimit-Limit	3600	Requests allowed per hour.
X-RateLimit-Remaining	3412	Requests left in the current hour.
X-RateLimit-Reset	1789243200	UTC epoch, in seconds, when the limit resets.

Response headers

HTTP/1.1 200 OK
Content-Type: application/json
X-RateLimit-Limit: 3600
X-RateLimit-Remaining: 3412
X-RateLimit-Reset: 1789243200

When you hit the limit

Once X-RateLimit-Remaining reaches zero, calls return TOO_MANY_REQUESTS until the time in X-RateLimit-Reset.

application/json

{
  "success": 0,
  "error": "TOO_MANY_REQUESTS"
}

Wait for the reset

Retrying immediately keeps failing. Sleep until X-RateLimit-Reset, then resume at a lower rate.

Volume add-ons

If 3,600 requests per hour is not enough, add volume to your plan. The add-on replaces your hourly limit and shows in X-RateLimit-Limit right away. See Pricing.

Volume add-ons
Requests / hour	Price / month	Average rate
200,000	$50	1 every 18 ms
500,000	$120	1 every 7 ms
1,000,000	$220	1 every 4 ms
3,000,000	$600	1 every 1 ms

Polling guidance

Use the lists to find what changed. One call to /v3/events/inplay returns score and timer for every live event. Fetch details only for the events you follow.
Fetch odds incrementally. On /v2/event/odds, pass since_time with the last add_time you stored and odds_market with the keys you use.
Walk history by day. The pager stops at page 100. On /v3/events/ended, use day instead of paging further.
Stop at final statuses. When an event's time_status is final, fetch it once more and stop polling. See time_status.
Back off on errors. Wait for the reset on TOO_MANY_REQUESTS, and retry INTERNAL_SERVER_ERROR with exponential backoff.