Need help to increase the Rate Limit for tool invocation #6449
Replies: 2 comments
|
Hi @stpforotx-ui, The error "Maximum 500 requests per minute for LOW tier endpoints" is coming from ContextForge's built-in rate limiting middleware — not an upstream provider. The tiers are assigned by path prefix:
Tool invocations should be hitting MEDIUM (100 rpm), not LOW (500 rpm). If you're seeing the LOW tier message it's worth double-checking which paths your client is calling. See the RFC 6585 compliance architecture docs for full details on how Error 429 responses are structured. How to raise the limits All limits are configurable via environment variables — no code change needed. Add these to your # MEDIUM tier — tool/MCP/LLM endpoints (default: 100 rpm)
RATE_LIMIT_MEDIUM_RPM=1000
RATE_LIMIT_MEDIUM_BURST=100
# LOW tier — health/metrics/docs (default: 500 rpm)
RATE_LIMIT_LOW_RPM=2000
RATE_LIMIT_LOW_BURST=200See Rate Limiting Configuration options for further options. To disable rate limiting entirely (e.g. if you manage it upstream at a load balancer or API gateway): RATE_LIMITING_ENABLED=falseI hope this helps, let me know how you get on. Thanks, |
|
Thank you this is very helpful. I was able to go past this once I changed it. LOW tier — health/metrics/docs (default: 500 rpm)RATE_LIMIT_LOW_RPM=2000 Could you please recommend. Thanks for all your help once again. |
Uh oh!
There was an error while loading. Please reload this page.
Hello Team,
We are using only a single user to invoke various tools across different MCP Server.
We are running into an issue - "Maximum 500 requests per minute for LOW tier endpoints"
Is there a way to increase the number of requests per minute.
Thanks for all your help in advance
All reactions