generated: '2026-08-28'
method: searched
source: https://freetts.org/developers
also_read:
- https://freetts.org/pricing
- https://freetts.org/llms.txt
probed: '2026-08-28'
limit_count: 5
exhaustion_status: 429
exhaustion_body: 'Too many requests. Please wait a minute.'
reset_behavior: fixed 60-second window; documented as resetting per IP, per minute
response_headers:
  observed: []
  note: >-
    PROBED, NOT ASSUMED. A live unauthenticated GET of
    https://freetts.org/api/voices on 2026-08-28 returned HTTP 200 with no
    RateLimit-*, no X-RateLimit-*, and no Retry-After header. The full response
    header set was date, content-type, content-length, server (cloudflare),
    x-frame-options, x-content-type-options, referrer-policy,
    strict-transport-security, permissions-policy, cf-cache-status, report-to,
    nel, cf-ray, alt-svc. An agent calling this API has no runtime budget signal
    at all - it can only discover the limit by being refused with a 429, and the
    429 body is prose rather than a structured retry hint.
  documented: []
limits:
- scope: per-IP (unauthenticated / documented on the API reference header block)
  window: 1 minute
  limit: 20
  source: https://freetts.org/developers
  note: >-
    The "Rate Limits" section of the developer page states "20 Requests per
    minute per IP".
- scope: per-API-key (free plan)
  window: 1 minute
  limit: 10
  source: https://freetts.org/developers
  note: >-
    The /developers FAQ ("Do I need an API key?" and "What happens when I hit
    the rate limit?") both state 10 requests per minute for a free API key. This
    contradicts the 20/min figure in the same page's header block, and the
    pricing page's feature matrix states a third figure, 15 req/min, for API
    access on the Free plan. All three are published by FreeTTS.
- scope: per-account (web tool, free plan)
  window: 1 minute
  limit: 15
  source: https://freetts.org/pricing
- scope: per-API-key (PRO)
  window: 1 minute
  limit: 200
  source: https://freetts.org/pricing
- scope: per-API-key (Creator)
  window: 1 minute
  limit: 1000
  source: https://freetts.org/pricing
quota_limits:
- scope: characters per request
  free: 1000
  pro: 10000
  creator: 25000
  source: https://freetts.org/developers (POST /tts parameter table) and llms.txt
  note: >-
    Contradicted by the same page's FAQ and by the pricing page, both of which
    say 5,000 characters per request on the free tier. The Chrome extension
    1.5.3 release notes describe fixing exactly this 1,000-vs-5,000 gap in the
    extension, which suggests 5,000 is the current server behaviour and the
    1,000 in the reference table is stale.
- scope: characters per month
  free: 15000
  pro: 1000000
  creator: 5000000
  source: https://freetts.org/pricing
  note: >-
    The /developers FAQ states 5,000 chars/month for a free API key, against
    15,000 on the pricing page. Unresolved.
usage_signal:
  documented: true
  fields:
  - chars_used
  - chars_limit
  source: >-
    /developers FAQ "What is the monthly character allowance?" states every
    generate call returns chars_used and chars_limit in the response body. Not
    verifiable here - POST /api/v1/tts requires a key and returns 401
    {"detail":"Missing x-api-key header"} without one.
findings:
- >-
  The published rate limit for the free tier takes three different values
  (10/min, 15/min, 20/min) across two pages of the provider's own
  documentation, and the per-request character cap takes two (1,000 and 5,000).
  An integrator cannot size a retry policy from these docs.
- >-
  No rate-limit response headers are emitted, so the numbers above are the only
  signal available and none of them can be confirmed at runtime.