generated: '2026-08-12'
method: searched
source: https://docs.ploy.ai/cli/remote-development
docs:
- https://docs.ploy.ai/cli/remote-development
- https://docs.ploy.ai/cli/authentication
- https://docs.ploy.ai/webhooks
limit_count: 2
rate_limits:
- name: API token requests
  scope: per-api-token
  limit: 60
  window: 1 minute
  burst: null
  status_on_exhaustion: null
  cli_exit_code: 1
  headers: []
  note: >-
    "Stay below 60 requests per minute for each API token." A rate-limited CLI
    command exits 1 and includes the retry delay in the message. Ploy does not
    document the HTTP status or any RateLimit-* / Retry-After response header for
    this surface, and there is no public REST reference where one could be read —
    an agent gets the retry delay only as prose in the CLI message.
  source: https://docs.ploy.ai/cli/remote-development
- name: Webhook ingest
  scope: per-endpoint
  limit: null
  window: null
  burst: null
  status_on_exhaustion: 429
  headers: []
  note: >-
    The webhook ingest endpoint returns 429 Too Many Requests when "the endpoint
    is being rate limited", but Ploy publishes no number, no window, and no
    Retry-After header contract. Documented guidance is sender-side: back off and
    retry on 5xx and 429 with jitter.
  source: https://docs.ploy.ai/webhooks
quotas:
- name: Webhook payload size
  limit: 1 MB per request
  status_on_exhaustion: 413
  source: https://docs.ploy.ai/webhooks
- name: Credits
  note: >-
    Ploy's primary consumption control is credits, not request rate — 800/day
    (3,000 max) on Free, 4,000/mo on Starter, 24,000/mo on Pro, custom on
    Enterprise. See plans/ploy-plans-pricing.yml.
  source: https://ploy.ai/pricing
- name: Visitor enrichment
  limit: 50/mo Starter, 1,000/mo Pro, custom Enterprise
  source: https://ploy.ai/pricing
response_headers:
  documented: false
  note: >-
    No X-RateLimit-*, RateLimit-* or Retry-After header is documented on any Ploy
    surface. This is the single biggest runtime-signal gap for an agent driving
    Ploy: the limit is known (60/min) but there is no header telling a caller how
    much of it is left.