Guides
Rate limits
40 requests an hour per key, 2 seconds apart, and the headers that tell you where you stand.
The limits
Each key gets 40 requests per clock hour and must leave 2 seconds between calls. Every endpoint counts toward the same budget. A brand that needs more can ask Growth Dept for a higher hourly limit.
Headers
| Header | Meaning |
|---|---|
X-RateLimit-Limit | The hourly limit |
X-RateLimit-Remaining | Requests left this hour |
X-RateLimit-Reset | When the hour's count starts over, in Unix seconds |
When a call is refused
A refused call is a 429 with Retry-After in seconds and a body like this. Calling sooner than 2 seconds after the last call gets Retry-After: 2.
{
"error": "Hourly rate limit of 40 requests exceeded.",
"retry_after": 1260,
"reset_time": "2026-11-04T15:00:00Z"
}Exceptions
POST /campaign_applicationsskips the spacing, because sign-up forms can submit at the same moment. It still counts toward the hour.GET /posts(the posting report) counts toward neither, so schedulers can poll it.
Staying under
Ask for bigger pages, cache what rarely changes (brands, campaigns, filters), and honour Retry-After. The MCP server has its own budget: 60 units a minute and 1,000 an hour per key.