Limits
Rate limits
| Scope | Default | Note |
|---|---|---|
| Per API key | 60 requests per minute | Applies to all /v1 endpoints. |
| Per tenant | 300 requests per minute | Adds up all the tenant's keys; creating more keys does not increase the total. |
Per end user (end_user) | 10 questions per minute | Only on /v1/ask; the 429 carries details.scope = "end_user". |
When a limit is exceeded, the response is 429 rate_limited with the Retry-After header (seconds). If your integration needs more, contact contato@tisolutio.com.br.
A real person does not come close to 10 questions per minute. If the end-user limit is reached often, the most likely cause is an end_user shared among several people or an automation.
Sizes
| Item | Limit |
|---|---|
| Request body | 96 KiB |
question | 2,000 characters |
max_rows | from 1 to 1,000 (default 100) |
Feedback comment | 500 characters |
History q | 200 characters |
| History page | 50 items |
Quantities per tenant
| Item | Limit |
|---|---|
| Active API keys | 20 |
| Question → SQL examples | 200 (all of them go into every prompt) |
| Evaluation cases | 50 |
Billing
There is no question quota and no credits. The monthly fee is based on the bands of active end users in the month, counted by the end_user field (see End-user identifier). LLM tokens are billed directly by your provider, on your account.