rate_limits:
  description: MonkeyLearn enforces three tiers of rate limiting to ensure fair usage across all API consumers. A "query" is the unit of consumption — one request may count as zero, one, or multiple queries depending on text count and model type. SDKs include built-in throttle handling with automatic retry logic.
  reference: https://help.monkeylearn.com/en/articles/2174290-does-monkeylearn-have-a-throttle-limit

  limits:
    - name: Monthly Query Limit
      description: Maximum number of queries allowed per billing month, determined by your plan tier. Exceeding this limit raises a PlanQueryLimitError.
      scope: account
      period: month
      unit: queries
      tiers:
        - plan: Free
          limit: 300
        - plan: Team
          limit: 100000
        - plan: Business
          limit: custom

    - name: Per-Minute Rate Limit
      description: Maximum number of API requests allowed per minute. SDK clients automatically retry with backoff when this limit is reached.
      scope: account
      period: minute
      unit: requests
      tiers:
        - plan: Free
          limit: 20
        - plan: Team
          limit: 120
        - plan: Business
          limit: custom

    - name: Per-Request Batch Size
      description: Maximum number of text items that can be submitted in a single API request. SDK auto-batching handles splitting larger datasets automatically.
      scope: request
      unit: texts
      limit: 200

  error_codes:
    - code: PlanQueryLimitError
      description: Monthly query quota for the current plan has been exceeded.
    - code: PlanRateLimitError
      description: Per-minute request rate limit has been exceeded; retry after a short delay.
    - code: ConcurrencyRateLimitError
      description: Too many simultaneous requests; reduce concurrency and retry.

  notes:
    - The MonkeyLearn Python, Ruby, Node.js, PHP, and Java SDKs include automatic throttle handling with configurable retry behavior.
    - Auto-batching (enabled by default) transparently splits large text arrays into compliant batches.
    - MonkeyLearn is now part of Medallia; enterprise rate limits are negotiated as part of the EDR pricing model.