Name: Incredibly Fast Whisper API: Speech to Text in 100 Languages
Brand: magicapi
SKU: api-magicapi-ultra-fast-speech-to-text-api-vaibhavs10-incredibly-fast-whisper
Availability: InStock
Author: API Market

Shashank Agarwal

# 🎧 Speech-to-Text Transcription API (vaibhavs10 Incredibly Fast Whisper)

## Overview
Incredibly Fast Whisper runs OpenAI's whisper-large-v3 speech recognition model with Hugging Face Transformers, set up for speed by its author (vaibhavs10).

Send an `audio` file URL. The response contains the full transcript in `text` and a list of `chunks`, each with its text and start and end time in seconds. Set `timestamp` to `word` for word-level timestamps, `task` to `translate` for an English translation, and `language` to skip language detection. You pay per request, by processing time.

## Features
- **Transcription in many languages**: `language` lists 100 languages; the default `None` detects the language automatically.
- **Translation**: set `task` to `translate` to get English text from speech in another language.
- **Timestamps**: `chunk` (default) gives segment timings; `word` gives a timing for every word.
- **Speed**: the example's 48.5-second clip was transcribed in 2.41 seconds of processing.
- **Batch size**: `batch_size` (default 24) sets how many audio batches are computed in parallel.
- **Speaker diarisation**: `diarise_audio` labels speakers with Pyannote.audio; it needs your own Hugging Face token in `hf_token`.

## Use Cases
1. **Meeting and interview transcripts**: searchable text with segment timings.
2. **Subtitles**: word or chunk timestamps for captions.
3. **Podcast and video indexing**: full text for search and summaries.
4. **Translation of recordings**: English text from speech in other languages.

## Visual Examples
### Example Input
Audio: [the example speech clip (WAV)](https://blog.api.market/wp-content/uploads/2026/10/incredibly-fast-whisper-input-api-market.wav)

Sent with `task` `transcribe` and `batch_size` `64`; `language` and `timestamp` were left at `None` and `chunk`. The request also sent `return_timestamps` `true`, which is not a field of this model's input schema.

### Example Output
```json
{
  "chunks": [
    {
      "text": " the little tales they tell are false the door was barred locked and bolted as well ripe pears are fit hours fly by much too soon. The room was crowded",
      "timestamp": [
        0,
        29.72
      ]
    },
    {
      "text": " with a mild wab. The room was crowded with a wild mob. This strong arm shall shield your",
      "timestamp": [
        29.72,
        38.98
      ]
    },
    {
      "text": " honour. She blushed when he gave her a white orchid The beetle droned in the hot June sun",
      "timestamp": [
        38.98,
        48.52
      ]
    }
  ],
  "text": " the little tales they tell are false the door was barred locked and bolted as well ripe pears are fit hours fly by much too soon. The room was crowded with a mild wab. The room was crowded with a wild mob. This strong arm shall shield your honour. She blushed when he gave her a white orchid The beetle droned in the hot June sun"
}
```

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `audio` | string (URL) | Yes |  |  | Audio file URL |
| `task` | string | No | `transcribe` | `transcribe`, `translate` | Task to perform: transcribe, or translate the speech into English |
| `language` | string | No | `None` | `None`, `afrikaans`, `albanian`, `amharic`, `arabic`, `armenian`, `assamese`, `azerbaijani`, `bashkir`, `basque`, `belarusian`, `bengali`, `bosnian`, `breton`, `bulgarian`, `cantonese`, `catalan`, `chinese`, `croatian`, `czech`, `danish`, `dutch`, `english`, `estonian`, `faroese`, `finnish`, `french`, `galician`, `georgian`, `german`, `greek`, `gujarati`, `haitian creole`, `hausa`, `hawaiian`, `hebrew`, `hindi`, `hungarian`, `icelandic`, `indonesian`, `italian`, `japanese`, `javanese`, `kannada`, `kazakh`, `khmer`, `korean`, `lao`, `latin`, `latvian`, `lingala`, `lithuanian`, `luxembourgish`, `macedonian`, `malagasy`, `malay`, `malayalam`, `maltese`, `maori`, `marathi`, `mongolian`, `myanmar`, `nepali`, `norwegian`, `nynorsk`, `occitan`, `pashto`, `persian`, `polish`, `portuguese`, `punjabi`, `romanian`, `russian`, `sanskrit`, `serbian`, `shona`, `sindhi`, `sinhala`, `slovak`, `slovenian`, `somali`, `spanish`, `sundanese`, `swahili`, `swedish`, `tagalog`, `tajik`, `tamil`, `tatar`, `telugu`, `thai`, `tibetan`, `turkish`, `turkmen`, `ukrainian`, `urdu`, `uzbek`, `vietnamese`, `welsh`, `yiddish`, `yoruba` | Language spoken in the audio. Use 'None' to detect the language automatically |
| `batch_size` | integer | No | `24` |  | Number of parallel batches to compute. Lower it if a request fails with an out-of-memory error |
| `timestamp` | string | No | `chunk` | `chunk`, `word` | Timestamp level: chunk (per segment) or word (per word) |
| `diarise_audio` | boolean | No | `false` |  | Diarise the audio (label speakers) with Pyannote.audio. Requires hf_token |
| `hf_token` | string | No | none |  | Your Hugging Face access token, used by Pyannote.audio for diarisation. The token's account must have accepted the terms of the pyannote/speaker-diarization-3.1 and pyannote/segmentation-3.0 models |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.0046 per API unit.
- **1 API unit = 1 request covering up to 4 seconds of processing.** A request that processes for longer uses one unit per started 4-second block (for example 9 seconds = 3 units). Longer audio takes longer to process and so uses more units.
- Processing time is the model's run time; time spent waiting in the queue is not billed.
- **Example cost**: the example request (a 48.5-second clip) processed for 2.41 seconds = 1 API unit = **$0.0046**.
- **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/ultra-fast-speech-to-text-api-vaibhavs10-incredibly-fast-whisper`

### 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/ultra-fast-speech-to-text-api-vaibhavs10-incredibly-fast-whisper/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "3ab86df6c8f54c11309d4d1f930ac292bad43ace52d10c80d87eb258b3c9f79c",
    "input": {
      "task": "transcribe",
      "audio": "https://blog.api.market/wp-content/uploads/2026/10/incredibly-fast-whisper-input-api-market.wav",
      "batch_size": 64,
      "return_timestamps": true
    }
  }'
```

Response (201 Created):
```json
{
  "id": "z488a5sp7drmy0d1c6pb11qcd4",
  "version": "3ab86df6c8f54c11309d4d1f930ac292bad43ace52d10c80d87eb258b3c9f79c",
  "status": "starting",
  "created_at": "2026-10-22T03:33:58.637Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/ultra-fast-speech-to-text-api-vaibhavs10-incredibly-fast-whisper/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "z488a5sp7drmy0d1c6pb11qcd4",
  "status": "succeeded",
  "output": {
    "chunks": [
      {
        "text": " the little tales they tell are false the door was barred locked and bolted as well ripe pears are fit hours fly by much too soon. The room was crowded",
        "timestamp": [
          0,
          29.72
        ]
      },
      {
        "text": " with a mild wab. The room was crowded with a wild mob. This strong arm shall shield your",
        "timestamp": [
          29.72,
          38.98
        ]
      },
      {
        "text": " honour. She blushed when he gave her a white orchid The beetle droned in the hot June sun",
        "timestamp": [
          38.98,
          48.52
        ]
      }
    ],
    "text": " the little tales they tell are false the door was barred locked and bolted as well ripe pears are fit hours fly by much too soon. The room was crowded with a mild wab. The room was crowded with a wild mob. This strong arm shall shield your honour. She blushed when he gave her a white orchid The beetle droned in the hot June sun"
  },
  "error": null,
  "created_at": "2026-10-22T03:33:58.637Z",
  "started_at": "2026-10-22T03:33:59.637Z",
  "completed_at": "2026-10-22T03:34:02.047Z"
}
```

`output` is an object: `text` is the full transcript and `chunks` is a list of segments, each with `text` and a `timestamp` pair of start and end seconds. With `timestamp` set to `word`, each chunk is a single word.

### 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 `audio` or a `language` value 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/ultra-fast-speech-to-text-api-vaibhavs10-incredibly-fast-whisper";

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: "3ab86df6c8f54c11309d4d1f930ac292bad43ace52d10c80d87eb258b3c9f79c", 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({
  "task": "transcribe",
  "audio": "https://blog.api.market/wp-content/uploads/2026/10/incredibly-fast-whisper-input-api-market.wav",
  "batch_size": 64,
  "return_timestamps": 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/ultra-fast-speech-to-text-api-vaibhavs10-incredibly-fast-whisper"


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": "3ab86df6c8f54c11309d4d1f930ac292bad43ace52d10c80d87eb258b3c9f79c", "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({
    "task": "transcribe",
    "audio": "https://blog.api.market/wp-content/uploads/2026/10/incredibly-fast-whisper-input-api-market.wav",
    "batch_size": 64,
    "return_timestamps": True
})
print(output)
```

## FAQ
- **How is a request priced?**
  - By processing time: $0.0046 for every started 4 seconds of processing. The example's 48.5-second clip took 2.41 seconds, so it cost $0.0046.
- **How do I get word-level timestamps?**
  - Set `timestamp` to `word`. Each item in `chunks` is then one word with its start and end time.
- **What does translate do?**
  - With `task` set to `translate`, the speech is translated and the output text is in English.
- **How do I label speakers?**
  - Set `diarise_audio` to `true` and pass your own Hugging Face access token in `hf_token`. The account behind the token must have accepted the terms of the pyannote/speaker-diarization-3.1 and pyannote/segmentation-3.0 models.
- **A long file failed with an out-of-memory error. What can I do?**
  - Lower `batch_size` (the default is 24) and try again.
- **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`.

## Contact Us
If you have any questions or need assistance, you can reach out to us through the embedded chat on the bottom right corner of the page or send an email to hello@magicapi.com.


Incredibly Fast Whisper API: transcribe or translate audio with Whisper large-v3, with chunk or word timestamps and optional speaker diarisation.