---
title: "Generate MP3 narration with a catalogue voice"
method: POST
path: "/v1/audio/speech"
---

# Generate MP3 narration with a catalogue voice

`POST /v1/audio/speech`

## Request

- Base URL: `https://api.neurodynamic.tech`
- URL: `https://api.neurodynamic.tech/v1/audio/speech`
- Auth: none required

Introductory pricing: 0.005 USDC for one MP3 up to 1,000 input characters. Seven catalogue voices; no account or voice approval. Read https://api.neurodynamic.tech/narration for request format and payment recovery.

## Headers

- `PAYMENT-SIGNATURE` string

## Request body

- object
  - `input` string, required
  - `voice` 'af_heart' | 'af_bella' | 'bf_emma' | 'bf_alice' | 'bm_george' | 'am_michael' | 'am_fenrir'
  - `model` 'kokoro'
  - `response_format` 'mp3'
  - `speed` unknown
  - `stream` unknown

## Response `200`

MP3 and PAYMENT-RESPONSE receipt

## Other responses

- `402`: Free x402 quote
- `409`: Payment conflict or uncertainty: do not pay again
- `503`: Busy: retry same proof

---

[API](https://skmtc.dev/neurodynamic/apis/neurodynamic-audio-and-llm-apis-narration-transcription-voic.md) · [All operations](https://skmtc.dev/neurodynamic/apis/neurodynamic-audio-and-llm-apis-narration-transcription-voic/llms.txt) · [OpenAPI document](https://skmtc.dev/neurodynamic/apis/neurodynamic-audio-and-llm-apis-narration-transcription-voic/revisions/b4363a913180?raw)