Core Concepts
Rate Limits
Each key or connected user can make 120 requests a minute.
Each API key, and each person using a connected AI app, gets 120 requests in any rolling 60 seconds, across every endpoint. Every successful response tells you where you stand.
| Header | Meaning |
|---|---|
X-RateLimit-Limit | Requests allowed per minute |
X-RateLimit-Remaining | Requests left in the current window |
Retry-After | On a 429, seconds to wait before trying again |
{
"error": {
"code": "rate_limited",
"message": "Rate limit of 120 requests per minute exceeded"
}
}Staying Under the Limit
- Back off when
X-RateLimit-Remaininggets low, and always waitRetry-Afterseconds after a429. - Use
limit=100when walking lists, so you make fewer requests. - Cache workspace IDs and profiles; they rarely change.
