Name: VieNeu-TTS API: Vietnamese Text to Speech and Voice Cloning
Brand: magicapi
SKU: api-magicapi-vietnamese-text-to-speech-api-nghimestudio-vieneu-tts
Availability: InStock
Author: API Market

Shashank Agarwal

# 🔊 Vietnamese Text-to-Speech API (nghimestudio VieNeu-TTS)

## Overview
VieNeu-TTS v3 Turbo from nghimestudio is a Vietnamese text-to-speech model. Its developer describes it as producing 48 kHz audio with 25 preset voices, instant voice cloning and emotion cues.

Send Vietnamese (or mixed Vietnamese and English) text in `text` and pick one of the preset voices in `voice`, or send a short reference recording in `ref_audio` to speak in that voice instead. Each request returns the URL of one WAV file.

## Features
- **25 preset voices**: choose one in `voice` (default `Hải Đăng`).
- **Instant voice cloning**: send a 3 to 8 second reference clip in `ref_audio`; it overrides the preset voice. `denoise` (default `true`) cleans and trims the clip first.
- **Emotion tags**: add `[cười]` (laugh), `[thở dài]` (sigh) or `[hắng giọng]` (clear throat) inside the text.
- **Bilingual text**: the input can mix Vietnamese and English.
- **Long text**: the text is synthesized in chunks of up to `max_chars` characters (32 to 512, default 256).
- **Sampling controls**: `temperature` (0.1 to 1.5, default 0.8), `top_p` (0.1 to 1, default 0.95), `top_k` (1 to 100, default 25) and `repetition_penalty` (1 to 2, default 1.2).

## Use Cases
1. **Vietnamese voice-overs**: narration for videos, e-learning and presentations.
2. **Audio content**: read articles, stories and notices aloud.
3. **Apps and assistants**: spoken replies and announcements in Vietnamese.
4. **Custom voices**: speak in a voice you have the rights to by sending a short sample in `ref_audio`.

## Visual Examples
### Example Input
One Vietnamese sentence ("Hello, this is a sample voice clip created by a text-to-speech service.") read with the preset voice `Hải Đăng` and the default settings.

```json
{
  "text": "Xin chào, đây là một đoạn giọng nói mẫu được tạo bởi dịch vụ chuyển văn bản thành giọng nói.",
  "top_k": 25,
  "top_p": 0.95,
  "voice": "Hải Đăng",
  "denoise": true,
  "max_chars": 256,
  "temperature": 0.8,
  "repetition_penalty": 1.2
}
```

### Example Output
[Listen to the example output](https://blog.api.market/wp-content/uploads/2026/10/vieneu-tts-output-ks5kvdec-api-market.wav) (WAV audio).

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `text` | string | Yes |  |  | Vietnamese (or bilingual) text to synthesize. Supports the emotion tags `[cười]`, `[thở dài]` and `[hắng giọng]` |
| `voice` | string | No | `Hải Đăng` | `Hải Đăng`, `Adam bựa`, `Trúc Ly`, `Thiện Minh`, `Mai Anh`, `Thùy Dung`, `Thiền Tâm Đức`, `Ngọc Huyền`, `Quang Sơn`, `Ngọc Trân`, `Minh Đức`, `Phạm Tuyên`, `Thái Sơn`, `Xuân Vĩnh`, `Thanh Bình`, `Ngọc Linh`, `Đoan Trang`, `Thục Đoan`, `Minh Triết`, `Mỹ Duyên`, `Quỳnh Anh`, `Đức Trí`, `Kim Thanh`, `Adam`, `Quốc Tuấn` | Built-in preset voice. Ignored when `ref_audio` is provided |
| `ref_audio` | string (URL) | No | none |  | Optional 3 to 8 second reference clip (URL) for instant voice cloning. When set, it overrides the preset voice |
| `denoise` | boolean | No | `true` |  | Denoise and trim the reference clip before cloning |
| `temperature` | number | No | `0.8` | `0.1` to `1.5` | Sampling temperature |
| `top_k` | integer | No | `25` | `1` to `100` | Top-k sampling |
| `top_p` | number | No | `0.95` | `0.1` to `1` | Top-p (nucleus) sampling |
| `repetition_penalty` | number | No | `1.2` | `1` to `2` | Repetition penalty |
| `max_chars` | integer | No | `256` | `32` to `512` | Maximum characters per synthesis chunk |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.001 per API unit.
- **1 API unit = 1 request with up to 9 seconds of processing.** A request that takes longer bills one more unit for each started block of 9 seconds (for example, 10 seconds of processing is 2 units).
- Longer texts take longer to synthesize, so a long text can use more than one unit.
- **Example cost**: the example request took 2.92 seconds of processing, which is 1 API unit = **$0.001**.
- **Failed or cancelled predictions are not billed.**
- **GET requests to check prediction status** are free and don't count toward API unit usage.
- **No monthly subscription fee.**

## API Documentation

### Base URL
All API requests go to: `https://prod.api.market/api/v1/magicapi/vietnamese-text-to-speech-api-nghimestudio-vieneu-tts`

### Authentication
Send your api.market API key in the `x-api-market-key` header:
```
x-api-market-key: YOUR_API_MARKET_KEY
```

### Asynchronous Processing
**IMPORTANT**: This API is asynchronous. The flow is:
1. Send a POST request to `/predictions`. A successful request returns **201 Created** with a prediction `id` and a `status` of `starting` or `processing`.
2. Poll GET `/predictions/{id}` with that ID.
3. Keep polling while `status` is `starting` or `processing`.
4. Stop when `status` is `succeeded` (read `output`), `failed` or `canceled` (read `error`).

A polling interval of 1 second and an overall timeout of 600 seconds (10 minutes) work well.

### Endpoints

#### Create a Prediction
```bash
curl -X POST "https://prod.api.market/api/v1/magicapi/vietnamese-text-to-speech-api-nghimestudio-vieneu-tts/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "7f9fe07b44b5cdf98a0a50152d3548122e812b1dc0c2c4baade48b63bb0bc041",
    "input": {
      "text": "Xin chào, đây là một đoạn giọng nói mẫu được tạo bởi dịch vụ chuyển văn bản thành giọng nói.",
      "top_k": 25,
      "top_p": 0.95,
      "voice": "Hải Đăng",
      "denoise": true,
      "max_chars": 256,
      "temperature": 0.8,
      "repetition_penalty": 1.2
    }
  }'
```

Response (201 Created):
```json
{
  "id": "ks5kvdecs5rmt0d1d21sd1nz8c",
  "version": "7f9fe07b44b5cdf98a0a50152d3548122e812b1dc0c2c4baade48b63bb0bc041",
  "status": "starting",
  "created_at": "2026-10-23T11:27:12.582Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/vietnamese-text-to-speech-api-nghimestudio-vieneu-tts/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "ks5kvdecs5rmt0d1d21sd1nz8c",
  "status": "succeeded",
  "output": "https://blog.api.market/wp-content/uploads/2026/10/vieneu-tts-output-ks5kvdec-api-market.wav",
  "error": null,
  "created_at": "2026-10-23T11:27:12.582Z",
  "started_at": "2026-10-23T11:27:13.082Z",
  "completed_at": "2026-10-23T11:27:16.004Z"
}
```

`output` is a single URL string pointing to the generated WAV file.

### Status Codes
- **201 Created**: prediction created (POST `/predictions`). This is the normal success response.
- **200 OK**: prediction details returned (GET `/predictions/{id}`).
- **400 Bad Request**: invalid input, for example a missing `text` or a `voice` that is not in the list.
- **401 Unauthorized**: missing or invalid API key.
- **404 Not Found**: no prediction with that ID.
- **429 Too Many Requests**: rate limit exceeded.
- **500 Internal Server Error** or **503 Service Unavailable**: temporary server problem; retry later.

## Rate Limits
All plans allow 10 requests per second. Requests above that return `429 Too Many Requests`; back off before retrying and honour any `Retry-After` header. Polling a prediction with GET is free but still counts toward the per-second limit, so poll every 1 to 5 seconds rather than in a tight loop.

## Error Handling
- Check the HTTP status code of every request and read the response body for the error message.
- A prediction can also end with `status` set to `failed`; the `error` field then explains why. Failed predictions are not billed.
- Retry 429, 500 and 503 responses with exponential backoff. Fix the input before retrying a 400.

## Implementation Guide
### Example Code

#### JavaScript
```javascript
const apiKey = "YOUR_API_MARKET_KEY";
const baseUrl = "https://prod.api.market/api/v1/magicapi/vietnamese-text-to-speech-api-nghimestudio-vieneu-tts";

async function createAndWaitForPrediction(input) {
  const response = await fetch(`${baseUrl}/predictions`, {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "x-api-market-key": apiKey
    },
    body: JSON.stringify({ version: "7f9fe07b44b5cdf98a0a50152d3548122e812b1dc0c2c4baade48b63bb0bc041", input })
  });

  if (response.status !== 201) {
    throw new Error(`API error: ${response.status} ${await response.text()}`);
  }

  const prediction = await response.json();
  return pollForResults(prediction.id);
}

const delay = (ms) => new Promise((resolve) => setTimeout(resolve, ms));

async function pollForResults(predictionId) {
  const maxTimeout = 600000;
  const pollInterval = 1000;
  const startTime = Date.now();

  while (Date.now() - startTime < maxTimeout) {
    const response = await fetch(`${baseUrl}/predictions/${predictionId}`, {
      headers: { "x-api-market-key": apiKey }
    });

    if (!response.ok) {
      throw new Error(`API error: ${response.status} ${await response.text()}`);
    }

    const prediction = await response.json();

    if (prediction.status === "succeeded") {
      return prediction.output;
    }

    if (prediction.status === "failed" || prediction.status === "canceled") {
      throw new Error(`Prediction ${prediction.status}: ${prediction.error || "Unknown error"}`);
    }

    await delay(pollInterval);
  }

  throw new Error("Prediction timed out after 10 minutes");
}

createAndWaitForPrediction({
  "text": "Xin chào, đây là một đoạn giọng nói mẫu được tạo bởi dịch vụ chuyển văn bản thành giọng nói.",
  "top_k": 25,
  "top_p": 0.95,
  "voice": "Hải Đăng",
  "denoise": true,
  "max_chars": 256,
  "temperature": 0.8,
  "repetition_penalty": 1.2
}).then((output) => console.log(output));
```

#### Python
```python
import time
import requests

API_KEY = "YOUR_API_MARKET_KEY"
BASE_URL = "https://prod.api.market/api/v1/magicapi/vietnamese-text-to-speech-api-nghimestudio-vieneu-tts"


def create_and_wait_for_prediction(input_data):
    headers = {"Content-Type": "application/json", "x-api-market-key": API_KEY}
    response = requests.post(
        f"{BASE_URL}/predictions",
        headers=headers,
        json={"version": "7f9fe07b44b5cdf98a0a50152d3548122e812b1dc0c2c4baade48b63bb0bc041", "input": input_data},
    )

    if response.status_code != 201:
        raise Exception(f"API error: {response.status_code} {response.text}")

    return poll_for_results(response.json()["id"])


def poll_for_results(prediction_id):
    max_timeout = 600
    poll_interval = 1
    headers = {"x-api-market-key": API_KEY}
    start_time = time.time()

    while time.time() - start_time < max_timeout:
        response = requests.get(f"{BASE_URL}/predictions/{prediction_id}", headers=headers)

        if response.status_code != 200:
            raise Exception(f"API error: {response.status_code} {response.text}")

        prediction = response.json()

        if prediction["status"] == "succeeded":
            return prediction["output"]

        if prediction["status"] in ("failed", "canceled"):
            raise Exception(f"Prediction {prediction['status']}: {prediction.get('error') or 'Unknown error'}")

        time.sleep(poll_interval)

    raise Exception("Prediction timed out after 10 minutes")


output = create_and_wait_for_prediction({
    "text": "Xin chào, đây là một đoạn giọng nói mẫu được tạo bởi dịch vụ chuyển văn bản thành giọng nói.",
    "top_k": 25,
    "top_p": 0.95,
    "voice": "Hải Đăng",
    "denoise": True,
    "max_chars": 256,
    "temperature": 0.8,
    "repetition_penalty": 1.2
})
print(output)
```

## FAQ
- **Which voices are available?**
  - `Hải Đăng`, `Adam bựa`, `Trúc Ly`, `Thiện Minh`, `Mai Anh`, `Thùy Dung`, `Thiền Tâm Đức`, `Ngọc Huyền`, `Quang Sơn`, `Ngọc Trân`, `Minh Đức`, `Phạm Tuyên`, `Thái Sơn`, `Xuân Vĩnh`, `Thanh Bình`, `Ngọc Linh`, `Đoan Trang`, `Thục Đoan`, `Minh Triết`, `Mỹ Duyên`, `Quỳnh Anh`, `Đức Trí`, `Kim Thanh`, `Adam`, `Quốc Tuấn`. The default is `Hải Đăng`.
- **How do I clone a voice?**
  - Send a 3 to 8 second recording of the voice in `ref_audio`. It overrides `voice`. Leave `denoise` on unless the clip is already clean. Only clone voices you have permission to use.
- **Can the text include English?**
  - Yes. The model accepts Vietnamese or bilingual text.
- **What format is the output?**
  - A URL to a WAV file. Download it soon after the prediction finishes and store your own copy.
- **How much does a request cost?**
  - $0.001 per request for up to 9 seconds of processing. The example took 2.92 seconds and cost $0.001.
- **What if my prediction fails?**
  - Failed predictions are not billed. Read the `error` field, adjust the input and try again.

## Contact Us
If you have any questions, reach out to us through the chat on the bottom right of the page. You can also email our support team at hello@magicapi.com.


Generate Vietnamese speech with VieNeu-TTS v3 Turbo: 25 preset voices, instant voice cloning from a 3 to 8 second clip and emotion tags.