Whisper is a set of multi-lingual, robust speech recognition models trained by OpenAI that achieve state-of-the-art results in many languages. Whisper models were trained to predict approximate timestamps on speech segments (most of the time with 1-second accuracy), but they cannot originally predict word timestamps. This version has implementation to predict word timestamps and provide a more accurate estimation of speech segments when transcribing with Whisper models.

You can use cURL or any other http client to run inferences:

```bash
curl -X POST \
    -H "Authorization: bearer $(deepctl auth token)"  \
    -F audio=@my_voice.mp3  \
    'https://api.deepinfra.com/v1/inference/openai/whisper-timestamped-medium'
```

which will give you back something similar to:

```json
{
  "text": "",
  "segments": [
    {
      "id": 0,
      "text": "Hello",
      "start": 0.0,
      "end": 1.0
    },
    {
      "id": 1,
      "text": "World",
      "start": 4.0,
      "end": 5.0
    }
  ],
  "language": "en",
  "request_id": null,
  "inference_status": {
    "status": "unknown",
    "runtime_ms": 0,
    "cost": 0.0,
    "tokens_generated": 0,
    "tokens_input": 0
  }
}

```


You can use our command-line tool [deepctl](/docs/getting-started) to run
inferences:

```bash
deepctl infer \
    -m 'openai/whisper-timestamped-medium'  \
    -i audio=@my_voice.mp3
```

which will give you back something similar to:

```json
{
  "text": "",
  "segments": [
    {
      "id": 0,
      "text": "Hello",
      "start": 0.0,
      "end": 1.0
    },
    {
      "id": 1,
      "text": "World",
      "start": 4.0,
      "end": 5.0
    }
  ],
  "language": "en",
  "request_id": null,
  "inference_status": {
    "status": "unknown",
    "runtime_ms": 0,
    "cost": 0.0,
    "tokens_generated": 0,
    "tokens_input": 0
  }
}

```


audio

task

language that the audio is in; uses detected language if None

language

temperature

patience

suppress_tokens

optional text to provide as a prompt for the first window.

initial_prompt

provide the previous output of the model as a prompt for the next window

condition_on_previous_text

temperature to increase when falling back when the decoding fails to meet either of the thresholds below

temperature_increment_on_fallback

compression_ratio_threshold

logprob_threshold

probability of the <|nospeech|> token threshold

no_speech_threshold

The webhook to call when inference is done, by default you will get the output in the response of your inference request

webhook

Audio

Compression Ratio Threshold

Condition On Previous Text

Initial Prompt

Language

Logprob Threshold

No Speech Threshold

Patience

Suppress Tokens

Task

Temperature

Temperature Increment On Fallback

Webhook

AutomaticSpeechRecognitionIn

Avg Logprob

Compression Ratio

confidence of the segment (Only in whisper-timestamped model)

Confidence

end location in input in seconds from start

No Speech Prob

Seek

start location in input in seconds from start

Start

Text

Tokens

a list of timestamped words in a segment (Only in whisper-timestamped model)

Words

Segment

estimated cost billed for the request in USD

Cost

Runtime Ms

Status

Tokens Generated

Tokens Input

Word

Object containing the status of the inference request

openai/whisper-timestamped-medium

HTTP/cURL API

Input fields

`audio`string

`task`string

`language`string

`temperature`number

`patience`number

`suppress_tokens`string

`initial_prompt`string

`condition_on_previous_text`boolean

`temperature_increment_on_fallback`number

`compression_ratio_threshold`number

`logprob_threshold`number

`no_speech_threshold`number

`webhook`file

Input Schema

Output Schema

openai/whisper-timestamped-medium

HTTP/cURL API

Input fields

audiostring

taskstring

languagestring

temperaturenumber

patiencenumber

suppress_tokensstring

initial_promptstring

condition_on_previous_textboolean

temperature_increment_on_fallbacknumber

compression_ratio_thresholdnumber

logprob_thresholdnumber

no_speech_thresholdnumber

webhookfile