Whisper V3 Large

State-of-the-art automatic speech recognition for transcribing audio to text.

Whisper V3 Large is OpenAI’s state-of-the-art automatic speech recognition model that transcribes audio to text. It supports multiple languages and can handle various audio formats with high accuracy.

Try in playground

Test Whisper V3 Large in the Runpod Hub playground.

Endpoint	https://api.runpod.ai/v2/whisper-v3-large/runsync
Pricing	$0.05 per 1000 characters
Type	Audio transcription

Request

All parameters are passed within the input object in the request body.

input.audio · string · required

URL of the audio file to transcribe.

input.prompt · string

Optional context or prompt to guide transcription. Useful for domain-specific terminology or formatting hints.

cURL

curl -X POST "https://api.runpod.ai/v2/whisper-v3-large/runsync" \
  -H "Authorization: Bearer $RUNPOD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "prompt": "",
      "audio": "https://example.com/audio-file.mp3"
    }
  }'

import requests

response = requests.post(
    "https://api.runpod.ai/v2/whisper-v3-large/runsync",
    headers={
        "Authorization": f"Bearer {RUNPOD_API_KEY}",
        "Content-Type": "application/json",
    },
    json={
        "input": {
            "prompt": "",
            "audio": "https://example.com/audio-file.mp3",
        }
    },
)

result = response.json()
print(result["output"]["transcription"])

const response = await fetch(
  "https://api.runpod.ai/v2/whisper-v3-large/runsync",
  {
    method: "POST",
    headers: {
      Authorization: `Bearer ${RUNPOD_API_KEY}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      input: {
        prompt: "",
        audio: "https://example.com/audio-file.mp3",
      },
    }),
  }
);

const result = await response.json();
console.log(result.output.transcription);

Response

id · string

Unique identifier for the request.

status · string

Request status. Returns COMPLETED on success, FAILED on error.

delayTime · integer

Time in milliseconds the request spent in queue before processing began.

executionTime · integer

Time in milliseconds the model took to transcribe the audio.

workerId · string

Identifier of the worker that processed the request.

output · object

The transcription result.

output.transcription · string

The transcribed text from the audio.

output.cost · float

Cost of the transcription in USD.

200

{
  "id": "sync-a1b2c3d4-e5f6-7890-abcd-ef1234567890-u1",
  "status": "COMPLETED",
  "delayTime": 18,
  "executionTime": 12345,
  "workerId": "oqk7ao1uomckye",
  "output": {
    "transcription": "Hello and welcome to this demonstration of Whisper V3. This is a sample transcription that shows the high accuracy of the model.",
    "cost": 0.0065
  }
}

{
  "id": "sync-a1b2c3d4-e5f6-7890-abcd-ef1234567890-u1",
  "status": "FAILED",
  "error": "Invalid audio URL or unsupported format"
}

Supported audio formats

Whisper V3 Large supports common audio formats including:

• MP3
• WAV
• FLAC
• M4A
• OGG

Cost calculation

Whisper V3 Large charges $0.05 per 1000 characters of transcribed audio. Example costs:

Characters	Cost
500 characters	$0.025
1,000 characters	$0.05
10,000 characters	$0.50