Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.critique-labs.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

Rate limits apply per API key and depend on your plan. Limits are enforced across Critique Flow API endpoints (including POST /v1/search and GET /v1/ws/search).
These limits are not currently defined in the published OpenAPI document. They are documented here and on the pricing page. When limits are added to the OpenAPI spec, this page will stay in sync with info.description and endpoint 429 responses.

​
Limits by plan

PlanRate limit
Free (Starter)10 API calls per minute
Pro100 API calls per minute
Enterprise5,000+ API calls per minute (custom)

​
Free tier during peak demand

The free tier may be throttled below the stated limit during peak usage so we can prioritize paid tier traffic. This is temporary and depends on overall platform load; you may see slower responses or stricter rate limiting even when you are under 10 requests per minute. We reserve the right to throttle the free tier when needed to support paid tier usage, paid tier usage itself however will never be subject to throttling.

​
Billing and tier behavior

  • Free tier: Applies when you are on the Starter plan (no paid usage). The 10 requests per minute limit applies.
  • Pro tier: 100 requests per minute when you have an active Pro plan or paid usage as described on pricing.
  • Token usage: There is no token cap on the free tier; large queries may take longer to process.

​
Fair use and abuse prevention

Creating multiple accounts (or otherwise splitting traffic across identities) to circumvent rate limits is prohibited. If we determine that accounts are being used to bypass limits, we reserve the right to suspend or permanently ban those accounts, with or without prior notice.

​
When you are rate limited

If you exceed your limit, the API may return an error such as: 
Rate limited, please try again in a minute
Implement exponential backoff, respect Retry-After when present, and cache responses where possible. See Search for endpoint details.