generated: '2026-08-13'
method: searched
source: https://app.usenotch.ai/mcp/setup
limit_count: 0
notes: >-
  Notch publishes no request-rate limits for its MCP surface. The setup guide and
  FAQs document a credit-based quota model instead of a rate limit: "Ads you create
  over MCP spend the same monthly credits as ads created on usenotch.ai, so your
  Notch plan is the only thing that matters." No per-minute/per-hour ceilings, no
  burst allowance and no exhaustion status code are documented anywhere on the
  setup guide, the pricing page or llms.txt. No rate-limit response headers were
  observed on the anonymous MCP probe (the endpoint answers 401 before any limit
  accounting), so the runtime signal an agent would need is not available.
limits: []
headers_observed: []
documented_headers: []
quota:
  model: credits
  scope: workspace/month
  reset: monthly
  shared_with: usenotch.ai web app (same credit pool)
  overage_behavior: undocumented
x-evidence:
- fetched: '2026-08-13'
  url: https://app.usenotch.ai/mcp/setup
  http_status: 200
  note: FAQ documents credits, not rate limits
- fetched: '2026-08-13'
  url: https://app.usenotch.ai/mcp
  http_status: 401
  note: 'anonymous POST tools/list; response carried no X-RateLimit-*, RateLimit-* or Retry-After header'
checked: '2026-08-13'