Limits
These are the limits during the developer beta. They can change.
Rate limits
Section titled “Rate limits”| What | Limit | Beyond it |
|---|---|---|
Token requests (POST …/connect/token) |
120 a minute per client address | 429 Too Many Requests |
Sign-in, sign-up and password forms on auth.nimles.com |
20 posts a minute per client address | 429 Too Many Requests |
| Creating your own tenants | 3 requests an hour per user | 429 Too Many Requests |
The limits are counted per server instance, so treat them as the most you can rely on, not as a guarantee of throughput. A token lives an hour: cache it, and the token limit is never in your way.
The API calls themselves (api.nimles.com) have no published rate limit during the beta. Be reasonable: run requests
for one tenant a few at a time, not hundreds in parallel. The MCP server allows 600 requests a minute per
client address.
Tenants
Section titled “Tenants”Each user can create 3 tenants of their own. Creating a fourth answers 409 Conflict
(You already have 3 tenants of your own.). Creating tenants needs developer-beta access (request it from
My tenants in the portal, see the quickstart) and a verified email address (403 otherwise).
Request size
Section titled “Request size”- A request body can be about 30 MB at most.
- Media files are uploaded as JSON: a
data:URI with the file in base64 inuri(POST /media/api/tenant/{tenantId}/Files). Base64 adds a third, so a file can be about 22 MB.
Timeouts
Section titled “Timeouts”A request that takes longer than 120 seconds is cut off by the gateway with 504 Gateway Timeout. No regular
call comes close; split large imports into smaller requests.
REST lists are not paged and answer every matching item. For large collections, use GraphQL, which pages at up to 100 items.