name: Keboola Rate Limits and Platform Limits
description: Keboola enforces both business (soft/contractual) limits tied to subscription
  tier and technical (hard) platform limits. API rate limits are not publicly documented
  with specific numbers; platform limits focus on storage, compute, and concurrency.
url: https://help.keboola.com/management/project/limits/
created: '2026-06-13'
modified: '2026-06-13'
authentication:
  method: API Token (X-StorageApi-Token header)
  description: Most Keboola APIs require a Storage API token passed via the X-StorageApi-Token
    HTTP header. Tokens are generated within Keboola projects and scoped to specific
    permissions.
  header: X-StorageApi-Token
business_limits:
  description: Contractual soft limits that vary by subscription tier. Exceeding
    these triggers alerts or requires plan upgrades.
  limits:
    - name: Projects
      description: Number of Keboola workspaces (projects) allowed per organization.
      tier: Enterprise (negotiated)
    - name: Users
      description: Count of project users per project. Excludes support users and
        API tokens.
      tier: Enterprise (negotiated)
    - name: Project Power Units (PPUs)
      description: Total compute time credits available per billing period. Measured
        across all jobs including data source, destination, transformation, and orchestration
        jobs.
      tier: Free (60 min/month), Enterprise (negotiated)
    - name: Storage Size
      description: Sum of sizes of all tables in Table Storage per project. Aliases,
        linked buckets, and files do not count toward this limit.
      tier: Free (250 GB), Enterprise (negotiated)
platform_limits:
  description: Technical hard limits enforced by the platform for performance and
    stability.
  limits:
    - name: Tables per Bucket
      description: Soft limit of 200 tables per single storage bucket. Exceeding
        this may affect UI performance and query planning.
      value: 200 (recommended maximum)
      type: soft
    - name: Free Plan Compute
      description: Free plan provides 120 compute minutes in the first month, then
        60 minutes per month. Additional minutes are billed at $0.14 per minute.
      value: 120 minutes (month 1), 60 minutes/month thereafter
      overage: $0.14 per minute
      type: hard (jobs pause when exhausted)
api_limits:
  description: Specific per-endpoint rate limits are not publicly documented. The
    Keboola platform uses token-based authentication with per-project scoping. Developers
    should implement exponential backoff for 429 responses.
  notes:
    - API rate limits are enforced at the platform level but not published with specific
      thresholds
    - Recommended to use exponential backoff and retry logic for production integrations
    - Regional stacks operate independently; limits apply per stack
    - Storage API is the primary API subject to request throttling
regional_stacks:
  description: Keboola operates across multiple regional stacks, each an independent
    instance. Limits apply per stack.
  stacks:
    - name: US Virginia (AWS)
      url: https://connection.keboola.com
    - name: EU Frankfurt (AWS)
      url: https://connection.eu-central-1.keboola.com
    - name: EU Ireland (AWS)
      url: https://connection.north-europe.azure.keboola.com
    - name: Custom single-tenant
      url: https://connection.CUSTOMER_NAME.keboola.com