Name: MiniMax Speech-02-HD API: High-Fidelity Text to Speech
Brand: magicapi
SKU: api-magicapi-text-to-audio-high-fidelity-voice-synthesis-minimax-speech-02-hd
Availability: InStock
Author: API Market

Shashank Agarwal

# 🔊 Text-to-Speech API (MiniMax Speech-02-HD)

## Overview
Speech-02-HD is a text-to-speech model from MiniMax. Its developer describes it as offering voice synthesis, emotional expression and multilingual capabilities, optimized for high-fidelity applications like voiceovers and audiobooks.

Send up to 10,000 characters of `text` and pick a `voice_id`. Each request returns the URL of one audio file in the format you choose.

## Features
- **Voices**: any MiniMax system voice through `voice_id` (default `English_Wiselady`; the example uses `Friendly_Person`).
- **Emotion**: `auto` (default), `happy`, `sad`, `angry`, `fearful`, `disgusted`, `surprised`, `calm`, `fluent` or `neutral`.
- **Pauses**: insert markers like `<#0.5#>` in the text to pause for that many seconds.
- **Delivery**: `speed` (0.5 to 2), `volume` (0 to 10) and `pitch` (-12 to +12 semitones).
- **Audio output**: `mp3` (default), `wav`, `flac` or `pcm`; sample rates from 8000 to 44100 Hz; MP3 bitrates from 32000 to 256000 bps; mono or stereo.
- **Language hint**: `language_boost` accepts `Automatic` or one of 41 languages and dialects, including English, Chinese, Cantonese, Spanish, French, German, Japanese, Korean, Arabic and Hindi.
- **English normalization**: `english_normalization` improves how numbers and dates are read.

## Use Cases
1. **Voiceovers**: narration for product videos, explainers and ads.
2. **Audiobooks and articles**: long-form narration in chunks of up to 10,000 characters.
3. **Multilingual audio**: the same script voiced in several languages with `language_boost`.
4. **Phone and app prompts**: short voice prompts, with low sample rates such as 8000 Hz for telephony.

## Visual Examples
### Example Input
Text (315 characters):

```text
Speech-02-series is a Text-to-Audio and voice cloning technology that offers voice synthesis, emotional expression, and multilingual capabilities.

The HD version is optimized for high-fidelity applications like voiceovers and audiobooks. While the turbo one is designed for real-time applications with low latency.
```

Settings: `voice_id` `Friendly_Person`, `emotion` `happy`, `language_boost` `English`, `english_normalization` `true`, `speed` 1, `volume` 1, `pitch` 0, `sample_rate` 32000, `bitrate` 128000, `channel` `mono`. `audio_format` was left at its default (`mp3`).

### Example Output
[Listen to the example output](https://blog.api.market/wp-content/uploads/2026/10/speech-02-hd-output-v8b483ya-api-market.mp3)

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `text` | string | Yes |  |  | Text to narrate (max 10,000 characters). Use markers like <#0.5#> to insert pauses in seconds. |
| `voice_id` | string | No | `English_Wiselady` |  | Voice to synthesize. Any MiniMax system voice (for example English_Wiselady or English_Deep-VoicedGentleman) or the voice_id of a MiniMax cloned voice. |
| `speed` | number | No | `1` | `0.5` to `2` | Speech speed multiplier (0.5 to 2.0). Lower is slower, higher is faster. |
| `volume` | number | No | `1` | `0` to `10` | Relative loudness. 1.0 is the default MiniMax gain. Range 0 to 10. |
| `pitch` | integer | No | `0` | `-12` to `12` | Semitone offset applied to the voice (-12 to +12). |
| `emotion` | string | No | `auto` | `auto`, `happy`, `sad`, `angry`, `fearful`, `disgusted`, `surprised`, `calm`, `fluent`, `neutral` | Desired delivery style. Use auto to let the model choose, or pick a specific emotion. |
| `english_normalization` | boolean | No | `false` |  | Improve number and date reading for English text (adds a small amount of latency). |
| `sample_rate` | integer | No | `32000` | `8000`, `16000`, `22050`, `24000`, `32000`, `44100` | Audio sample rate in Hz. |
| `bitrate` | integer | No | `128000` | `32000`, `64000`, `128000`, `256000` | MP3 bitrate in bits per second. Only used when audio_format is mp3. |
| `audio_format` | string | No | `mp3` | `mp3`, `wav`, `flac`, `pcm` | File format of the generated audio: mp3 for general use, wav or flac for lossless, or pcm for raw bytes. |
| `channel` | string | No | `mono` | `mono`, `stereo` | mono for 1 channel (default), stereo for 2 channels. |
| `subtitle_enable` | boolean | No | `false` |  | Request MiniMax subtitle metadata with sentence timestamps. |
| `language_boost` | string | No | `None` | `None`, `Automatic`, `Chinese`, `Chinese,Yue`, `Cantonese`, `English`, `Arabic`, `Russian`, `Spanish`, `French`, `Portuguese`, `German`, `Turkish`, `Dutch`, `Ukrainian`, `Vietnamese`, `Indonesian`, `Japanese`, `Italian`, `Korean`, `Thai`, `Polish`, `Romanian`, `Greek`, `Czech`, `Finnish`, `Hindi`, `Bulgarian`, `Danish`, `Hebrew`, `Malay`, `Persian`, `Slovak`, `Swedish`, `Croatian`, `Filipino`, `Hungarian`, `Norwegian`, `Slovenian`, `Catalan`, `Nynorsk`, `Tamil`, `Afrikaans` | Optional language hint. Choose Automatic to let the model detect the language, or pick a specific language. |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.013 per API unit.
- **1 API unit = 100 weighted tokens; an input token counts 1, an output token counts 0.** For this model 1 input token is 1 character of `text`.
- That is **$130 per million input tokens (characters)** and **$0 per million output tokens**, or $0.013 per 100 characters. Each request is rounded up to a whole API unit.
- Voice, emotion, speed, pitch and audio format settings do not change the price. The longest allowed text (10,000 characters) costs $1.30.
- **Example cost**: the example text is 315 characters (315 input tokens), which rounds up to 4 API units: **$0.052**.
- **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/text-to-audio-high-fidelity-voice-synthesis-minimax-speech-02-hd`

### 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. The example request finished in about 4 seconds.

### Endpoints

#### Create a Prediction
```bash
curl -X POST "https://prod.api.market/api/v1/magicapi/text-to-audio-high-fidelity-voice-synthesis-minimax-speech-02-hd/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "minimax/speech-02-hd",
    "input": {
      "text": "Speech-02-series is a Text-to-Audio and voice cloning technology that offers voice synthesis, emotional expression, and multilingual capabilities.\n\nThe HD version is optimized for high-fidelity applications like voiceovers and audiobooks. While the turbo one is designed for real-time applications with low latency.",
      "pitch": 0,
      "speed": 1,
      "volume": 1,
      "bitrate": 128000,
      "channel": "mono",
      "emotion": "happy",
      "voice_id": "Friendly_Person",
      "sample_rate": 32000,
      "language_boost": "English",
      "english_normalization": true
    }
  }'
```

Response (201 Created):
```json
{
  "id": "v8b483ya5xrmt0d1c71tgj6y0w",
  "version": "minimax/speech-02-hd",
  "status": "starting",
  "created_at": "2026-10-22T03:59:43.828Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/text-to-audio-high-fidelity-voice-synthesis-minimax-speech-02-hd/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "v8b483ya5xrmt0d1c71tgj6y0w",
  "status": "succeeded",
  "output": "https://blog.api.market/wp-content/uploads/2026/10/speech-02-hd-output-v8b483ya-api-market.mp3",
  "error": null,
  "created_at": "2026-10-22T03:59:43.828Z",
  "started_at": "2026-10-22T03:59:44.328Z",
  "completed_at": "2026-10-22T03:59:47.719Z"
}
```

`output` is a single URL string pointing to the generated audio file in the requested `audio_format`.

### 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 `speed` outside 0.5 to 2.
- **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/text-to-audio-high-fidelity-voice-synthesis-minimax-speech-02-hd";

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: "minimax/speech-02-hd", 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": "Speech-02-series is a Text-to-Audio and voice cloning technology that offers voice synthesis, emotional expression, and multilingual capabilities.\n\nThe HD version is optimized for high-fidelity applications like voiceovers and audiobooks. While the turbo one is designed for real-time applications with low latency.",
    "pitch": 0,
    "speed": 1,
    "volume": 1,
    "bitrate": 128000,
    "channel": "mono",
    "emotion": "happy",
    "voice_id": "Friendly_Person",
    "sample_rate": 32000,
    "language_boost": "English",
    "english_normalization": true
  }).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/text-to-audio-high-fidelity-voice-synthesis-minimax-speech-02-hd"


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": "minimax/speech-02-hd", "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")


print(create_and_wait_for_prediction({
    "text": "Speech-02-series is a Text-to-Audio and voice cloning technology that offers voice synthesis, emotional expression, and multilingual capabilities.\n\nThe HD version is optimized for high-fidelity applications like voiceovers and audiobooks. While the turbo one is designed for real-time applications with low latency.",
    "pitch": 0,
    "speed": 1,
    "volume": 1,
    "bitrate": 128000,
    "channel": "mono",
    "emotion": "happy",
    "voice_id": "Friendly_Person",
    "sample_rate": 32000,
    "language_boost": "English",
    "english_normalization": True
}))
```

## FAQ
- **How is my text counted?**
  - Each character of `text` is one input token. 100 characters make one API unit ($0.013), rounded up per request.
- **How do I add a pause?**
  - Put a marker like `<#0.5#>` in the text. The number is the pause length in seconds.
- **Which audio formats can I get?**
  - `mp3` (default), `wav`, `flac` or `pcm`, set with `audio_format`. `bitrate` only applies to MP3.
- **Does `subtitle_enable` change the response?**
  - The response `output` is the audio URL only; this API does not return a separate subtitle field.
- **Is the 201 response from POST an error?**
  - No. 201 Created means the prediction was accepted. Use the returned `id` to poll GET `/predictions/{id}` until `status` is `succeeded`, `failed` or `canceled`.
- **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 high-fidelity speech from up to 10,000 characters of text with MiniMax Speech-02-HD: voices, emotions, speed, pitch and language hints.