Substy Rate limits
API and MCP traffic share one request budget per agency.
Limits use your agency’s best active plan. Every API key and MCP connection for that agency shares the same budget across server instances.
| Agency plan | Calls per minute | Concurrent calls |
|---|---|---|
| Free | 20 | 2 |
| Pro | 60 | 5 |
| Elite | 120 | 10 |
Redis coordinates admission centrally. Creating another API key, switching clients, or using MCP does not create another allowance.
Inspect response headers
RateLimit-Limit: 60;w=60
X-RateLimit-Concurrency: 5When the agency exceeds a limit, the API returns 429 with Retry-After, expressed in seconds. Wait at least that long before retrying. Queue work and add randomized backoff so all workers do not restart together.
Creator platforms may enforce additional limits or conditions even when your agency has capacity. Read the error code and message rather than assuming every rejection is an agency rate limit.
Writes and retries
If a write may already have been accepted, inspect its operation receipt before submitting it again. See operations. Rate limiting does not make repeated writes safe.
Pricing during rollout
There is no separate metered API or MCP usage charge in this rollout. Existing platform actions can still have their normal costs. Usage measurement helps inform a future pricing strategy; the request limits above apply now.