> ## Documentation Index
> Fetch the complete documentation index at: https://platform.novisurf.top/llms.txt
> Use this file to discover all available pages before exploring further.

# TTS Generation

Learn how to instantly generate lifelike audio from text using Novisurf's fast TTS API.

## Overview

The Novisurf speech endpoint lets you convert text to spoken audio in seconds. It is fully OpenAI-compatible — just swap the base URL and API key and your existing code works out of the box.

## Endpoint

| Endpoint | Usage                 | URL                                              |
| -------- | --------------------- | ------------------------------------------------ |
| Speech   | Convert text to audio | `POST https://api2.novisurf.top/v1/audio/speech` |

## Authentication

**Bearer token (recommended)**

```
X-API-Key: lsk_...
```

**X-API-Key header**

```
Authorization: Bearer lsk_...
```

Your API key is available in the [Novisurf Dashboard](https://novisurf.top/dashboard).

## Supported Models

| Model ID          | Description                        |
| ----------------- | ---------------------------------- |
| `shellai/melotts` | Fast, lightweight multilingual TTS |

## Quick Start

The speech endpoint takes four key inputs:

* **model:** the TTS model to use
* **input:** the text to generate audio from
* **voice:** the desired voice for output
* **response\_format:** defaults to `mp3`

### Parameters

| Parameter         | Type   | Required | Description                                             |
| ----------------- | ------ | -------- | ------------------------------------------------------- |
| `model`           | string | Yes      | TTS model ID.                                           |
| `input`           | string | Yes      | Text to synthesize.                                     |
| `voice`           | string | Yes      | Voice to use for output.                                |
| `response_format` | string | No       | Output audio format. Defaults to `mp3`.                 |
| `speed`           | number | No       | Playback speed from `0.25` to `4.0`. Defaults to `1.0`. |

***

### Python

```python theme={null}
from openai import OpenAI

client = OpenAI(
    api_key="lsk_...",
    base_url="https://api2.novisurf.top/v1"
)

speech_file_path = "speech.mp3"

response = client.audio.speech.create(
    model="shellai/melotts",
    voice="en-us",
    input="Hello! Welcome to Novisurf. Fast inference, built for developers.",
    response_format="mp3"
)

response.stream_to_file(speech_file_path)
```

### JavaScript

```js theme={null}
import fs from "fs";
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: "lsk_...",
  baseURL: "https://api2.novisurf.top/v1",
});

const speechFilePath = "speech.mp3";

async function main() {
  const response = await client.audio.speech.create({
    model: "shellai/melotts",
    voice: "en-us",
    input: "Hello! Welcome to Novisurf. Fast inference, built for developers.",
    response_format: "mp3",
  });

  const buffer = Buffer.from(await response.arrayBuffer());
  await fs.promises.writeFile(speechFilePath, buffer);

  console.log(`Speech generated: ${speechFilePath}`);
}

main().catch(console.error);
```

### cURL

```bash theme={null}
curl https://api2.novisurf.top/v1/audio/speech \
  -X POST \
  -H "X-API-Key: lsk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "shellai/melotts",
    "input": "Hello! Welcome to Novisurf. Fast inference, built for developers.",
    "voice": "en-us",
    "response_format": "mp3"
  }' \
  --output speech.mp3
```

***

## Response

The response is raw audio binary in the requested format, not JSON. Save the response body directly to a file.

| Header         | Value                                     |
| -------------- | ----------------------------------------- |
| `Content-Type` | `audio/mpeg` for mp3, `audio/wav` for wav |

***

## Error Codes

| Status | Meaning                                     |
| ------ | ------------------------------------------- |
| `400`  | Bad request — missing or invalid parameters |
| `401`  | Unauthorized — invalid or missing API key   |
| `402`  | Insufficient credits                        |
| `429`  | Rate limit exceeded                         |
| `500`  | Internal server error                       |