Rate limits and usage
Limits apply per organization and per API, across every key you own. Past one, the API answers 429 Too Many Requests with a Retry-After header: wait that many seconds, then send the request again.
An agreement may set other figures. A second publishable key adds no quota: every key of your organization shares the same limits.
Only your own clients and the hosted embed count: requests made with your keys from your servers, pages, and SDK calls, the MCP server, and the tiles the hosted embed draws. They count toward your limits, your allowance, and your invoice.
An Embedded Studio, Map Studio in the Mountaya app, and share links are recorded in your usage but never counted. The Tile API's base map is never counted either.
The Data API and the Tile API report what is left of each limit on every answer, so you can slow down before a 429. Send a request with your keys and print the headers:
For example:
RateLimit-Policy names each limit with its quota q and window w in seconds. RateLimit gives, in the same order, the requests left r and the seconds until the window resets t.
Wait the seconds Retry-After names, plus a little random jitter, then send the same request. A write retried with its Idempotency-Key is never made twice.
A 429 for a spent monthly allowance names the seconds until your next billing period begins. Do not retry before then.
The Map Studio API limits reads per IP address or per member, then per organization. Exports, copies, and its costliest writes, such as imports, routed segments, and agent runs, have limits of their own per member. Imports also count the requests they send the routing engine, and agent runs the bytes they are sent.
POST /v1/sessions spends none of your quotas. It counts only refused requests, so a working backend never meets its limits, and up to 6,000 tokens a minute per secret key.
Open Organization → Usage in the Mountaya app to see what each API served this month, with the hosted embed, Embedded Studio, and share links shown apart.