Your account

Rate limits

Handle account and provider limits without retry storms.

OPENAI BASE URLhttps://www.modelmesh.in/api/v1

Account-level limits

The gateway checks requests and tokens per minute using the active plan. Limits apply to the user account; more API keys do not multiply capacity. The current implementation tracks rate counters in server memory.

Respond to HTTP 429

Wait for Retry-After when provided. Otherwise use capped exponential backoff with jitter. Reduce concurrency rather than immediately resending. Providers can impose independent limits.

Allowance is separate

HTTP 402 insufficient_quota is a funds or allowance problem, not a request-per-minute problem. Check Billing. Pricing shows active plan allowances and key limits.