generated: '2026-08-13'
method: searched
source:
- https://docs.dyspatch.io/api/
- openapi/_original/dyspatch-openapi-original.json
docs: https://docs.dyspatch.io/api/
limit_count: 0
note: >-
  Dyspatch publishes NO numeric rate limit — not a requests-per-second ceiling,
  not a per-plan quota, not a burst allowance. What it does publish is the
  runtime signal, and it publishes it well: every single response definition in
  the OpenAPI (success and error alike) declares an X-RateLimit-Remaining
  response header, and a dedicated 429 "Rate limit exceeded" response is wired
  onto every operation in the spec. An agent can therefore always see how much
  budget is left and can always detect exhaustion; it just cannot know the
  ceiling in advance. limit_count is an honest 0: the headers are documented, the
  numbers are not.
headers:
- name: X-RateLimit-Remaining
  description: The number of requests left for the time window.
  type: integer
  scope: unknown
  present_on: all responses
  evidence: >-
    Declared in components.responses on every shared response object
    (InternalError, InvalidRequest, RateLimited, Unauthenticated, Unauthorized,
    NotFound, Forbidden) in the published OpenAPI 3.0.0 document.
missing_headers:
- X-RateLimit-Limit
- X-RateLimit-Reset
- RateLimit
- RateLimit-Policy
- Retry-After
missing_headers_note: >-
  Neither the RFC 9910 (RateLimit / RateLimit-Policy) family nor the legacy
  X-RateLimit-Limit / X-RateLimit-Reset pair is documented, and no Retry-After is
  declared on the 429. A client that trips the limit is told that it tripped but
  not when it may retry, so back-off has to be guessed.
exhaustion:
  status: 429
  response_name: RateLimited
  description: Rate limit exceeded
  body_schema: APIError
  body_fields: [code, message, parameter]
  retry_after_header: false
  applies_to: every operation in the spec (109 paths, 131 operations)
limits: []
limits_note: >-
  No published limits. Searched the API reference (docs.dyspatch.io/api/), the
  API changelog, the implementation guide, the API key documentation, the
  knowledge-base llms.txt index, and the pricing page — none states a request
  ceiling, a window, a burst allowance, or a per-plan quota.
scopes_checked:
- scope: per-api-key
  documented: false
- scope: per-account
  documented: false
- scope: per-endpoint
  documented: false
window: unknown
window_note: >-
  The X-RateLimit-Remaining description says "for the time window" but the
  window length is never stated.
related:
  conventions: conventions/dyspatch-conventions.yml
  errors: errors/dyspatch-problem-types.yml
  plans: plans/dyspatch-plans-pricing.yml
x-evidence:
- url: https://docs.dyspatch.io/api/
  http_status: 200
  fetched: '2026-08-13'
- url: https://www.dyspatch.io/pricing/
  http_status: 200
  fetched: '2026-08-13'