specificationVersion: "0.1"
id: cognee-rate-limits
name: Cognee Rate Limits
description: >
  Rate limit information for the Cognee AI memory and knowledge graph REST API.
  Cloud-hosted limits are tied to the active subscription plan. Self-hosted
  (local Docker or private cloud) deployments have no imposed rate limits.
url: https://docs.cognee.ai/api-reference/introduction

rateLimits:
  - name: Free Plan API Calls
    description: >
      The Free tier does not publish a hard API call ceiling but access is intended
      for exploration and development use only. Heavy production workloads should
      upgrade to a paid plan.
    plan: Free
    type: requests
    limit: unspecified
    period: month
    scope: account

  - name: Developer Plan API Calls
    description: >
      Developer plan includes 10,000 API calls per month. Additional document
      storage and processing capacity can be added via top-up packs priced per
      document volume.
    plan: Developer
    type: requests
    limit: 10000
    period: month
    scope: account
    overagePolicy: >
      Additional capacity available via top-up packs: +1,000 docs for $35,
      +3,000 docs for $100, or +15,000 docs for $750.

  - name: Cloud (Team) Plan API Calls
    description: >
      Cloud Team plan includes 10,000 API calls per month shared across up to 10
      user accounts. Additional document capacity available via top-up packs.
    plan: Cloud (Team)
    type: requests
    limit: 10000
    period: month
    scope: account
    overagePolicy: >
      Additional capacity available via top-up packs: +1,000 docs for $35,
      +3,000 docs for $100, or +15,000 docs for $750.

  - name: Enterprise Plan API Calls
    description: >
      Enterprise (On-Prem) plans have custom rate limits negotiated based on
      deployment size, infrastructure, and contractual requirements.
    plan: On-Prem (Enterprise)
    type: requests
    limit: custom
    period: month
    scope: account

  - name: Self-Hosted (Local/Docker)
    description: >
      Self-hosted Cognee deployments running locally or in a private cloud impose
      no API rate limits. Throughput is constrained only by the underlying
      infrastructure resources.
    plan: Self-Hosted
    type: requests
    limit: unlimited
    period: none
    scope: instance

notes:
  - Limits are enforced at the account level, not per individual API key.
  - The API base URL for cloud is https://api.cognee.ai with a documented 99.9% uptime SLA.
  - For detailed or updated rate limit information contact support via Discord (https://discord.gg/m63hxKsp4p) or via the dedicated Slack channel on Team/Enterprise plans.