aid: respeecher
name: Respeecher Rate Limits
description: >-
  Documented rate and usage limits across Respeecher's two public APIs, captured
  from public documentation as of 2026-07-11. Values may change; confirm with
  the official docs.
source:
- url: https://docs.respeecher.com/
  type: Documentation
- url: https://space.respeecher.com/docs/limits
  type: Documentation
voice_marketplace_api:
  base_url: https://gateway.respeecher.com
  auth: API key via api-key header (create at marketplace.respeecher.com/account)
  limits:
  - name: Fast requests
    limit: 500 requests
    window: 300 seconds
  - name: Slow requests
    limit: 100 requests
    window: 300 seconds
  - name: Daily requests (plan-dependent)
    note: >-
      Subscription plans cap daily requests - roughly 350/day (TTS Only),
      100/day (Creator), 500/day (Power). Enterprise offers higher/unlimited
      concurrency.
space_realtime_tts_api:
  base_url: https://api.respeecher.com/v1/public/tts/en-rt
  auth: X-API-Key header, or api_key query-string parameter (for browser WebSocket clients)
  limits:
  - name: Concurrent generations
    limit: 5 concurrent text-to-speech generations per account across all endpoints
  - name: Bytes endpoint max response size
    limit: 100 MB per POST /tts/bytes response
  - name: Bytes endpoint transcript length
    limit: ~5,000 characters per POST /tts/bytes call
    note: >-
      The Bytes endpoint runs under a 120-second HTTP request timeout. For
      longer transcripts use the SSE or WebSocket streaming endpoints.
  - name: WebSocket connection lifetime
    limit: Connections not guaranteed to stay alive beyond 12 hours; idle connections may be closed server-side