---
title: "Synthesize speech for a text snippet — returns audio."
method: POST
path: "/voice/speak"
tags: ["voice"]
---

# Synthesize speech for a text snippet — returns audio.

`POST /voice/speak`

## Request

- Base URL: `https://api.livegraph.ai` (the document declares no server; this is the origin it was published from)
- URL: `https://api.livegraph.ai/voice/speak`
- Auth: one of:
  - API key in cookie `livegraph_session`
  - API key in header `x-api-key`

## Request body

- object
  - `text` string, required
  - `graphId` string, uuid
  - `voice` object
    - `provider` 'elevenlabs' | 'local' | 'openai-compatible' | 'openai', required
    - `voiceId` string, required

## Response `200`

Success — see summary for the response shape.

## Other responses

- `4XX`: Client error — body is the shared {error, code?} envelope.

---

[API](https://skmtc.dev/livegraph/apis/livegraph-api.md) · [All operations](https://skmtc.dev/livegraph/apis/livegraph-api/llms.txt) · [OpenAPI document](https://skmtc.dev/livegraph/apis/livegraph-api/revisions/551085d7719e?raw)