specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Dagger
providerId: dagger
generated: '2026-09-07'
method: searched
source: >-
  Searched https://docs.dagger.io/ via its published llms.txt index (no
  rate-limit, throttling or quota entry), https://docs.dagger.io/faq/ (no
  occurrence of "rate limit", "quota", "throttl" or "429"), and
  https://dagger.io/pricing (which meters plans but publishes no request limits).
created: '2026-05-04'
modified: '2026-09-07'
supersedes: >-
  This file previously held a bulk-sweep scaffold dated 2026-05-04 asserting
  X-RateLimit-* response headers, a 429 on exhaustion, and per-tier limits of
  10 rpm / 1,000 requests a month. NONE of that was published by Dagger — it was
  scaffold default text. Removed 2026-09-07 and replaced with the honest result.
limit_count: 0
description: >-
  Dagger publishes no rate limits, and for the engine that is structurally
  correct rather than an omission. The Dagger API is not a hosted multi-tenant
  service: it is a loopback endpoint at http://127.0.0.1:$DAGGER_SESSION_PORT/query
  served by an engine running on the caller's own machine for the life of one
  session. There is no shared capacity to protect and no other tenant to protect
  it from; the practical limits are the host's CPU, memory, disk and the engine's
  container concurrency.
limits: []
headers:
  limit: null
  remaining: null
  reset: null
  retry_after: null
  policy: null
  note: >-
    No rate-limit response headers are documented, and none were observed. The
    transport returns HTTP 200 for failed operations (GraphQL errors[] envelope),
    so there is no exhaustion status code to record either.
exhaustion_status: null
metered_surface:
  product: Dagger Cloud
  unit: Monthly Events
  note: >-
    The one place Dagger does meter is the hosted control plane, and it meters
    ingested telemetry volume rather than API calls: 1,000,000 monthly events on
    Individual, 10,000,000 on Team, "Contact us" on Enterprise
    (https://dagger.io/pricing). What happens on exceeding that is not published —
    no overage price, no throttle behaviour, no header. See
    plans/dagger-plans-pricing.yml.
finding: >-
  An agent integrating against Dagger Cloud has no runtime signal for how close it
  is to its event quota. The number exists on the pricing page; nothing surfaces
  it in a response.