Rate limits

Each limit counts requests in a fixed window. A per-minute limit resets at the start of every minute, not 60 seconds after your first request.

Every request

Caller	Limit
Organization API key	2,000 a minute per organization
Organization API key acting for a user with `TM-On-Behalf-Of`	Also 100 a minute for each user you act for
User token	100 a minute per user, and 1,000 a minute per IP address
No token	1,000 a minute per IP address
Any caller of `/v1/auth`	1,000 a minute per IP address, shared with the per-IP limit above

These endpoints also count against a stricter limit of their own.

Counted	Limit
User token, per user	10 a second and 120 a minute
Organization API key, per organization	100 a second and 1,200 a minute

Create new order: POST /v1/gateway/orders
Execute pending order: POST /v1/gateway/orders/{id}/execute
Prepare moving resting DeFi TP/SL orders to new trigger prices: POST /v1/gateway/orders/modify/{id}/prepare
Execute a prepared TP/SL modify: POST /v1/gateway/orders/modify/{id}/execute
Modify an active limit order: PATCH /v1/gateway/orders/{id}

Counted	Limit
User token, per user	120 a minute
Organization API key, per organization	1,200 a minute
No token, per IP address	120 a minute

Counted	Limit
User token, per user	20 a minute and 300 a day
Organization API key, per organization	200 a minute and 3,000 a day

Counted	Limit
User token, per user	10 a minute and 50 a day
Organization API key, per organization	100 a minute and 500 a day

Counted	Limit
Per IP address, for every caller	30 a minute

Counted	Limit
Per IP address, for every caller	5 a minute and 50 a day

We answer 429. A limit on this page answers with an empty body. A limit inside a service answers with the error body and type resource_exhausted.

Responses carry these headers for the limit closest to running out:

Header	Meaning
`X-RateLimit-Limit`	The limit and its window in seconds, such as `120, 120;w=60`.
`X-RateLimit-Remaining`	Requests left in the window.
`X-RateLimit-Reset`	Seconds until the window resets.

Wait for X-RateLimit-Reset, then retry. Your token stays valid. The headers can describe a limit that is not enforced yet.