generated: '2026-08-13'
method: searched
source: >-
  https://www.kana.ai (marketing site + FAQ + get-started), https://apps.kana.ai
  (application host), and the OpenAPI document Kana's console generates for a deployed
  skill (https://apps.kana.ai/index.js)
limit_count: 0
limits: []
response_headers: []
exhaustion_status: null
retry_after: false
notes: >-
  Kana publishes no rate limits. There is no developer portal, no public API reference
  and no limits table on any Kana surface. The contract Kana's own console generates
  for a deployed skill documents 200/401/500 only — no 429, no Retry-After, and no
  X-RateLimit-* or RateLimit-* response headers. The one throughput signal Kana does
  expose is queue_sz on GET /run/{runid}/status, which reports the platform-side job
  backlog; that is a capacity/latency hint, not a quota, and it is not recorded as a
  limit here. An honest zero.
x-evidence:
- url: https://www.kana.ai/docs
  status: 404
- url: https://www.kana.ai/developers
  status: 404
- url: https://apps.kana.ai/docs
  status: 401
- url: https://apps.kana.ai/index.js
  status: 200