Name: F5-TTS API: Zero-Shot Voice Cloning Text to Speech
Brand: magicapi
SKU: api-magicapi-voice-cloning-api-x-lance-f5-tts
Availability: InStock
Author: API Market

Shashank Agarwal

# 🗣️ Voice Cloning Text-to-Speech API (X-LANCE F5-TTS)

## Overview
F5-TTS from the X-LANCE lab is an open-source voice cloning text-to-speech model; its developer describes it as the new state of the art in open-source voice cloning. Send a short reference recording in `ref_audio` and the text to speak in `gen_text`, and the API returns a WAV file of that text spoken in the reference voice.

Add `ref_text`, the transcript of the reference clip, and adjust `speed`, `remove_silence` and `custom_split_words` as needed.

## Features
- **Voice cloning from a short clip**: the voice in `ref_audio` is used for the generated speech.
- **Any text**: put the words to speak in `gen_text`.
- **Speed control**: `speed` from 0.1 to 3 (default 1).
- **Silence removal**: `remove_silence` (default `true`) trims silences automatically.
- **Custom split words**: `custom_split_words` takes a comma-separated list of words used to split long text.
- **WAV output**: one audio file per request.

## Use Cases
1. **Narration in your own voice**: record a short sample once, then generate voice-overs from text.
2. **Product and app voices**: give prompts, notifications and tutorials a consistent voice.
3. **Prototyping audio content**: draft podcasts, ads or e-learning audio before recording in a studio.
4. **Accessibility**: read text aloud in a familiar voice.

Only clone voices you have the right to use, such as your own or a voice actor who has agreed to it.

## Visual Examples
### Example Input
Reference audio: [Listen to the example reference clip](https://blog.api.market/wp-content/uploads/2026/10/f5-tts-input-v2-api-market.wav)

Reference transcript (`ref_text`): "Hey there my name is Tara, and I'm a speech generation model that can sound like a person."

Text to speak (`gen_text`): "Welcome back. Today we will walk through three new features in our weekly product update."

Settings: `remove_silence` `true`, `custom_split_words` empty, `speed` left at its default of 1.

### Example Output
[Listen to the example output](https://blog.api.market/wp-content/uploads/2026/10/f5-tts-output-gkjj21tt-api-market.wav)

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `gen_text` | string | Yes |  |  | Text to generate |
| `ref_audio` | string (URL) | Yes |  |  | Reference audio for voice cloning |
| `ref_text` | string | No | none |  | Reference text: the transcript of `ref_audio` |
| `remove_silence` | boolean | No | `true` |  | Automatically remove silences |
| `custom_split_words` | string | No | `""` |  | Custom split words, comma separated |
| `speed` | number | No | `1` | 0.1 to 3 | Speed of the generated audio |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.012 per API unit.
- **1 API unit = 1 request with up to 7 seconds of processing.** A request that takes longer is billed one more unit for each started 7-second block (for example, 15 seconds of processing is 3 units, $0.036).
- Longer `gen_text` takes longer to generate and can add units.
- The example request (two sentences, 15 words) took 7.2 seconds of processing, so it was billed 2 units ($0.024).
- **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/voice-cloning-api-x-lance-f5-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
Replace `gen_text` with your own text, and `ref_audio` and `ref_text` with your own reference clip and its transcript.
```bash
curl -X POST "https://prod.api.market/api/v1/magicapi/voice-cloning-api-x-lance-f5-tts/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "87faf6dd7a692dd82043f662e76369cab126a2cf1937e25a9d41e0b834fd230e",
    "input": {
      "gen_text": "Welcome back. Today we will walk through three new features in our weekly product update.",
      "ref_text": "Hey there my name is Tara, and I'm a speech generation model that can sound like a person.",
      "ref_audio": "https://blog.api.market/wp-content/uploads/2026/10/f5-tts-input-v2-api-market.wav",
      "remove_silence": true,
      "custom_split_words": ""
    }
  }'
```

Response (201 Created):
```json
{
  "id": "gkjj21ttchrn80d1d9mbqe6qnc",
  "version": "87faf6dd7a692dd82043f662e76369cab126a2cf1937e25a9d41e0b834fd230e",
  "status": "starting",
  "created_at": "2026-10-23T20:16:26.394Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/voice-cloning-api-x-lance-f5-tts/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "gkjj21ttchrn80d1d9mbqe6qnc",
  "status": "succeeded",
  "output": "https://blog.api.market/wp-content/uploads/2026/10/f5-tts-output-gkjj21tt-api-market.wav",
  "error": null,
  "created_at": "2026-10-23T20:16:26.394Z",
  "started_at": "2026-10-23T20:16:27.394Z",
  "completed_at": "2026-10-23T20:16:35.394Z"
}
```

`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 `gen_text` or `ref_audio`, or a `speed` above 3.
- **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 (for example, a reference audio URL that cannot be downloaded). 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/voice-cloning-api-x-lance-f5-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: "87faf6dd7a692dd82043f662e76369cab126a2cf1937e25a9d41e0b834fd230e", 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({
  gen_text: "Welcome back. Today we will walk through three new features in our weekly product update.",
  ref_text: "Hey there my name is Tara, and I'm a speech generation model that can sound like a person.",
  ref_audio: "https://blog.api.market/wp-content/uploads/2026/10/f5-tts-input-v2-api-market.wav",
  remove_silence: true,
  custom_split_words: ""
}).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/voice-cloning-api-x-lance-f5-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": "87faf6dd7a692dd82043f662e76369cab126a2cf1937e25a9d41e0b834fd230e", "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({
    "gen_text": "Welcome back. Today we will walk through three new features in our weekly product update.",
    "ref_text": "Hey there my name is Tara, and I'm a speech generation model that can sound like a person.",
    "ref_audio": "https://blog.api.market/wp-content/uploads/2026/10/f5-tts-input-v2-api-market.wav",
    "remove_silence": True,
    "custom_split_words": "",
}))
```

## FAQ
- **What reference audio should I use?**
  - A short, clean recording of one speaker, with little background noise. Send its transcript in `ref_text`.
- **Can I change how fast the voice speaks?**
  - Yes. Set `speed` between 0.1 and 3 (default 1).
- **What does `custom_split_words` do?**
  - It takes a comma-separated list of words that the model uses as split points in the text. Leave it empty to use the defaults.
- **Can I clone any voice?**
  - Only clone voices you have permission to use, such as your own or a consenting speaker's.
- **What if my prediction fails?**
  - Failed predictions are not billed. Read the `error` field, check that `ref_audio` is reachable 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.


Clone a voice from a short reference clip with F5-TTS from X-LANCE and turn any text into speech in that voice, with speed control and silence removal.