---
title: "Generate speech"
method: POST
path: "/audio/speech"
tags: ["Audio"]
---

# Generate speech

`POST /audio/speech`

## Request body

- AudioSpeechRequest
  - `model` string, required — OpenAI TTS model — `openai/tts-1`, `openai/tts-1-hd`, `openai/gpt-4o-mini-tts`. Gemini TTS preview models are not served on this endpoint; use Gemini's native `/v1beta/` surface with `responseModalities: ["AUDIO"]` instead.
  - `input` string, required
  - `voice` string, required — Voice identifier. The gateway forwards this field to the upstream unchanged — use whichever voice name the upstream model supports.
  - `instructions` string — Optional voice direction (e.g. style, tone). Honored by models that support it, such as `openai/gpt-4o-mini-tts`.
  - `speed` number — Playback rate. Default `1.0`.
  - `stream_format` string — When set, response is streamed in this format chunk-by-chunk (subject to upstream support).
  - `response_format` 'mp3' | 'opus' | 'aac' | 'flac' | 'wav' | 'pcm'

## Response `200`

Binary audio content.

---

[API](https://skmtc.dev/orcarouter/apis/orcarouter-api.md) · [All operations](https://skmtc.dev/orcarouter/apis/orcarouter-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/orcarouter/orcarouter-api/revisions/9c5cc5d00311/schema)