generated: '2026-08-27'
method: searched
source: https://docs.shadowfax.ai/docs/introduction/plans-and-usage
docs: https://docs.shadowfax.ai/docs/introduction/plans-and-usage
limit_count: 0
note: >-
  An honest zero. Shadowfax AI names rate limits as a plan differentiator
  ("Standard rate limits" on Free, "Higher rate limits" on Pro) and the docs
  devote a "Current Usage Limits" section to the subject, but NOT ONE NUMERIC
  LIMIT IS PUBLISHED anywhere on the public surface. The docs describe the
  limits only qualitatively — "Generous limits", "No arbitrary row limits",
  "Fair usage limits on AI processing", "99% of users will never hit them" —
  and state that a user who approaches a limit will be contacted. There is
  also no public developer API, so there are no rate-limit response headers to
  observe: the only HTTP surface is app.shadowfax.ai, which requires a Clerk
  session JWT. Nothing here is fabricated to fill the artifact.
limits: []
headers:
  observed: false
  note: >-
    No response headers could be observed. app.shadowfax.ai answers unauthenticated
    requests with either a Clerk 401 or the SPA shell; no RateLimit-*, X-RateLimit-*
    or Retry-After header is documented or reachable.
  documented: []
exhaustion_status_code: null
qualitative_limits:
- scope: file-upload
  statement: 'Individual file uploads: "Generous limits for CSV, Excel, and Parquet files"'
- scope: workbook-data
  statement: 'Total data per Workbook: "Designed to handle typical analytical workloads"'
- scope: database-connections
  statement: '"No arbitrary row limits—query what you need"; connection limits based on fair usage'
- scope: views-per-workbook
  statement: Unlimited Views per Workbook
- scope: ai-processing
  statement: '"Fair usage limits on AI processing (LLM calls)"'
evidence:
- url: https://docs.shadowfax.ai/docs/introduction/plans-and-usage
  status: 200
- url: https://shadowfax.ai/pricing
  status: 200
- url: https://app.shadowfax.ai/openapi.json
  status: 401