v
latestOpenAPI 3.1.02026-08-0814845.6 KBOpenAI-compatible streaming speech
POST /v1/audio/speech with the response body chunked as it decodes. Point the standard OpenAI TTS client in LiveKit or Pipecat here to trade whole-utterance latency for first-sample latency. Both "input" and "text" are accepted.
Headers
Routing objective. Defaults to the route policy.
BCP 47 language tag used for benchmark selection and provider transcription.
Comma-separated providers or provider:model ids to allow.
Comma-separated providers or provider:model ids to exclude.
Maximum published benchmark price for the request stage: STT in USD per minute, LLM in USD per 1M tokens, or TTS in USD per 1M characters. Candidates without a published price are excluded when this constraint is set.
Request body
Response
Signed 16-bit mono 24 kHz PCM, chunked as each vendor chunk decodes. x-speko-first-byte-ms is absent by design: it is not yet known when these headers are written, which is the point of the route.