Rate Limits

Overview

Rate limits are enforced per API key at the Nginx gateway. Limits apply independently per region — a request to US and JP each count against their respective region limits.

Tier Limits

Tier	Requests/min	Requests/sec	Concurrent Streams
Basic	100	2	1
Pro	120,000	200	10
Quant	Unlimited	Unlimited	Unlimited

Protocol-Specific Limits

JSON-RPC

• Global Nginx limit: 5,000 req/s per API key
• Tier-specific limits apply within the global cap
• Batch requests count as one request but are limited to 100 calls per batch

gRPC

• Streaming RPCs count as one request at connection time
• Individual messages within a stream are not rate-limited
• Connection limits per tier apply

Unified Stream (SSE)

• SSE connections count against concurrent stream limits
• Event throughput is not rate-limited once connected
• Reconnect attempts count as new requests

Disk-Sync WebSocket

• Quant only — no rate limits
• One connection per API key

Rate Limit Headers

Rate-limited responses include:

Header	Description
X-RateLimit-Limit	Maximum requests per window
X-RateLimit-Remaining	Remaining requests in window
X-RateLimit-Reset	Unix timestamp when window resets
Retry-After	Seconds until next request allowed (on 429)

429 Response

{
  "error": "rate_limit_exceeded",
  "message": "Rate limit exceeded. Retry after 1 second.",
  "retry_after": 1
}

Best Practices

1. Use streaming protocols (gRPC, SSE, WebSocket) instead of polling
2. Implement exponential backoff on 429 responses
3. Cache responses client-side where appropriate
4. Use the region closest to your infrastructure to avoid wasted round-trips