Limits
How many calls a key and a workspace may make, how big a request and an answer can be, and what the 429 answer with Retry-After means.
Last updated 4 min read
On this page · 4 sections
| Limit | Current's default |
|---|---|
| All calls, per key | 120 a minute |
| Changes, per key | also 30 a minute |
| Changes to one ticket, per key | 20 an hour |
| All calls, per workspace | 50,000 a day (UTC) |
| Rows in a page | 100 at most |
| Request body | 64 KB at most |
| One answer | 256 KB, at most 15,000 characters in any one text |
Past a limit
The answer is 429 rate_limited with a Retry-After header: wait that many seconds, then try again. The API's edge at api.current.day also limits each key and each address before a call reaches Current, more generously than Current does, and answers the same way. The edge checks each request against the same rules as the API and never refuses a request the API would accept.
curl -sS --retry 3 --retry-all-errors "https://api.current.day/v1/boards" \
-H "Authorization: Bearer $CURRENT_API_KEY"
# curl --retry waits for Retry-After on a 429 by itselfLower the limits
A tenant admin can lower the workspace's limits, or one agent key's own, on the API access page, never raise them above Current's. A change applies to new calls within 30 seconds. These limits cover agent keys and connected apps; a personal MCP key someone made for their own AI app has its own limits, per person.
Long answers
An answer is never cut silently. A text longer than its limit is cut on a character boundary and its envelope says truncated with its full_length; a list that was cut is named in truncation; the answer itself says truncated.
When the page itself had to give up rows to fit, next_cursor is null and page_cut says how many rows were kept of how many. That null does not mean you have every row: ask again with the same cursor and limit set to the number kept.
What are the Current API's rate limits?
By default 120 calls a minute per key, of which at most 30 may be changes, 20 changes to one ticket an hour per key, and 50,000 calls a day per workspace. Past a limit the answer is 429 with Retry-After. A tenant admin can lower them on the API access page, and a change applies within 30 seconds. See the current-api-limits article. The Current API is not open yet.