generated: '2026-08-14'
method: searched
source: docs.photon.health (all 13 guide + reference pages) + live response headers
limit_count: 0
documented: false
note: >-
  Photon publishes no rate-limit policy. Every page listed in the docs llms.txt
  index was fetched as Markdown and searched for rate/limit/throttle/quota/429/
  Retry-After/X-RateLimit - zero matches across all of them. No rate-limiting
  response headers were observed on live unauthenticated responses from
  photonhealth.com or clinical-api.photon.health either. An honest zero: the
  limits may well exist at the gateway, but a consumer cannot discover them
  before hitting one.
headers_observed: []
documented_limits: []
exhaustion_behavior:
  status_code: 429
  method: derived
  confidence: low
  note: >-
    Not documented by Photon. errors/photon-error-codes.yml carries a rate_limit
    class inferred from ordinary gateway behavior; treat as a convention, not a
    published contract.
recommendation: >-
  Publish per-token limits and emit RateLimit-Limit / RateLimit-Remaining /
  RateLimit-Reset (RFC 9331) or X-RateLimit-* plus Retry-After on 429, so agents
  can back off on signal rather than on failure.
evidence:
- url: https://docs.photon.health/llms.txt
  status: 200
- url: https://docs.photon.health/docs/integration-guide.md
  status: 200
- url: https://docs.photon.health/reference/clinical-api.md
  status: 200
- url: https://clinical-api.photon.health/graphql
  status: 200