Rate Limits
BlueNexus applies rate limiting to protect the platform and ensure fair usage.
Limits by Endpoint
| Endpoint | Limit | Window |
|---|---|---|
| Global default | 100 requests | 1 minute |
POST /mcp |
30 requests | 1 minute |
POST /v1/agents/:id/chat/completions |
30 requests | 1 minute |
POST /v1/chat/completions |
30 requests | 1 minute |
POST /v1/auth/authenticate |
10 requests | 1 minute |
POST /v1/auth/token |
10 requests | 1 minute |
POST /v1/auth/revoke |
10 requests | 1 minute |
POST /v1/auth/register (DCR) |
5 requests | 1 minute |
Rate Limit Response
When you exceed a rate limit, the API returns 429 Too Many Requests:
{
"error": "TooManyRequestsError",
"message": "Rate limit exceeded",
"statusCode": 429
}
Credit Guard
In addition to rate limits, certain endpoints require a positive credit balance:
| Endpoint | Guard |
|---|---|
POST /mcp |
Credit balance must be > 0 |
POST /v1/agents/:id/chat/completions |
Credit balance must be > 0 |
POST /v1/chat/completions |
Credit balance must be > 0 |
If credits are exhausted, the API returns 402 Payment Required:
{
"error": "CreditsExhaustedError",
"message": "Insufficient credit balance",
"statusCode": 402
}
Best Practices
- Implement exponential backoff — On 429 responses, wait and retry with increasing delays.
- Batch where possible — Combine multiple tasks into a single
use-agentcall instead of making separate calls. - Cache responses — Cache provider listings and tool definitions that don't change frequently.
- Monitor usage — Track the
X-Credits-ConsumedandX-Credits-Remainingheaders to stay aware of your balance.