generated: '2026-09-03'
method: searched
source: https://www.getdietly.com/api-guide + https://www.getdietly.com/api + https://www.getdietly.com/developers/error-handling/
limit_count: 4
limits:
  - scope: per-IP (anonymous and Starter/free-key traffic)
    window: 1 minute
    limit: 30
    burst: null
    note: Free keys share the per-IP limit; an additional hourly anti-scraping cap applies; ~350ms fairness delay on anonymous/free reads.
  - scope: per-IP hourly anti-scraping cap (anonymous/Starter)
    window: 1 hour
    limit: null
    burst: null
    note: Documented as "an hourly cap that only a scraper would reach"; exact number not published.
  - scope: per-account (Pro plan, key sent)
    window: 1 minute
    limit: 500
    burst: null
  - scope: per-account (Scale plan, key sent)
    window: 1 minute
    limit: 3000
    burst: null
headers:
  - name: X-RateLimit-Limit
    meaning: The plan's requests-per-minute allowance (or the key's own cap, if lower).
  - name: X-RateLimit-Remaining
    meaning: Requests left in the current minute window (0 on a 429).
  - name: X-RateLimit-Reset
    meaning: Seconds until the window resets.
  - name: Retry-After
    meaning: Seconds to wait, sent on 429.
exhaustion:
  status: 429
  body: '{"detail": "Rate limit exceeded. Retry in N seconds."}'
  guidance: Wait for Retry-After, then capped exponential backoff; docs recommend pacing when X-RateLimit-Remaining drops below 5.
notes: >-
  Rate-limit headers are reported on authenticated responses; a live anonymous probe of GET /search
  (2026-09-03, HTTP 200) returned no X-RateLimit-* headers. Per-key rate caps can also be set in the
  dashboard. Enterprise limits are custom.