generated: '2026-09-19'
method: probed
source: https://fetch-price.com/docs/#limits
docs: https://fetch-price.com/docs/#limits
limit_count: 5
summary: >-
  One published request-rate limit and four plan quotas. Docs: "30 requests/minute per key (or per IP
  when keyless). Every response carries X-RateLimit-Limit and X-RateLimit-Remaining headers; excess
  requests get HTTP 429. Monthly lookup quotas are per plan." Both headers were OBSERVED live on
  2026-09-19 on every api.fetch-price.com response, including 400s and 404s: X-RateLimit-Limit 30,
  X-RateLimit-Remaining counting down 30 -> 29 -> 27 across consecutive unauthenticated requests from
  one IP. No 429 was provoked.
rate_limits:
- id: per-minute
  scope: per-key (per-IP when no key is sent)
  window: 1 minute
  limit: 30
  burst: null
  applies_to: all endpoints on https://api.fetch-price.com (headers seen on /, /health, /api/stats, /api/query, /a2a)
  source: docs + observed headers
- id: quota-free
  scope: per-key or per-IP
  window: 1 month
  limit: 50 lookups (POST /api/query)
  plan: Free
  source: docs
- id: quota-pro
  scope: per-key
  window: 1 month
  limit: 5000 lookups
  plan: Pro (GBP 29/mo)
  source: docs
- id: quota-scale
  scope: per-key
  window: 1 month
  limit: unlimited
  plan: Scale (GBP 99/mo)
  source: docs
- id: quota-trade
  scope: per-key
  window: 1 month
  limit: unlimited
  plan: Trade (GBP 99/mo)
  source: docs
signaling:
  scope: per-key / per-IP
  exhaustion_status: 429
  response_headers:
  - name: X-RateLimit-Limit
    documented: true
    observed: true
    observed_value: '30'
  - name: X-RateLimit-Remaining
    documented: true
    observed: true
    observed_value: '30, 29, 27 on successive requests'
  - name: X-RateLimit-Reset
    documented: false
    observed: false
  - name: Retry-After
    documented: false
    observed: false
  - name: RateLimit / RateLimit-Policy (IETF draft)
    documented: false
    observed: false
  client_guidance: >-
    Read X-RateLimit-Remaining on every response and pause before it reaches 0; on 429 wait for the
    next minute window (no reset time is signalled). A 429 can also mean the monthly quota is spent - the
    provider's MCP server surfaces that as "rate limit reached on current tier - see
    fetch-price.com/pricing"; the two causes are not distinguishable from the response.
  quota_visibility: >-
    No endpoint or header reports remaining monthly lookups; registration returns query_limit once.
domains:
- api.fetch-price.com