generated: '2026-09-21'
method: searched
source: https://openhandle.dev/pricing.md
name: Openhandle rate limits
limits:
- name: Free key throughput
  limit_text: 2 requests per second per key
  max: 2
  window: '1 second'
  scope: per-key
  plan: Free
- name: Pay-as-you-go key throughput
  limit_text: 10 requests per second per key
  max: 10
  window: '1 second'
  scope: per-key
  plan: Pay as you go
headers:
  policy: RateLimit-Policy
  runtime: RateLimit
  retry: Retry-After
  standard: draft-ietf-httpapi-ratelimit-headers-11 (Internet-Draft)
  exhaustion_status: 429
  exhaustion_code: RATE_LIMITED
scopes_enforced:
- key
- organization (workspace)
note: >-
  Openhandle enforces separate API key and workspace (organization) limits; Test
  and Live use separate limits, and Live trial limits can differ from paid. Every
  response carries RateLimit-Policy and RateLimit headers
  (draft-ietf-httpapi-ratelimit-headers-11, an Internet-Draft) whose values are
  snapshots, not reserved capacity — read them rather than assume a fixed
  allowance; the documented example policy is
  'RateLimit-Policy: "key";q=10;w=1, "organization";q=50;w=1' (q=allowance,
  w=window seconds, r=remaining, t=seconds to reset). On a limit, the API returns
  429 RATE_LIMITED with Retry-After; a limiter outage returns 503. The concrete
  per-second caps (2 rps Free, 10 rps paid, per key) are published on the pricing
  page; the quota-header shape is documented at
  https://openhandle.dev/docs/rate-limits. Volume customers get higher limits by
  negotiation. Failed requests are not billed.