Table of Contents

Method Transcribe

Namespace
LMKit.Speech
Assembly
LM-Kit.NET.dll

Transcribe(WaveFile, string, CancellationToken)

Synchronously transcribes the provided audio content into text segments.

public SpeechToText.TranscriptionResult Transcribe(WaveFile audioFile, string language = "auto", CancellationToken cancellationToken = default)

Parameters

audioFile WaveFile

The WAV-format audio content to transcribe.

language string

The language identifier to transcribe in, or "auto" to auto-detect.

cancellationToken CancellationToken

Token to cancel the operation.

Returns

SpeechToText.TranscriptionResult

A SpeechToText.TranscriptionResult containing transcription segments.

Examples

var model = LM.LoadFromModelID("whisper-large-turbo3");
var engine = new SpeechToText(model);
var audioFile = new WaveFile("audio.wav");
var result = engine.Transcribe(audioFile, language: "auto");
foreach (var segment in result.Segments)
    Console.WriteLine(segment.Text);

Remarks

Supported values:

  • "auto" to enable language auto-detection.
  • Any language identifier returned by GetSupportedLanguages() for the loaded model (typically ISO-639-1 codes such as "en", "fr", "de", "es").

Values are model-dependent; use GetSupportedLanguages() to query the exact list at runtime.

Exceptions

ArgumentNullException

Thrown if audioFile is null.

NotSupportedException

Thrown if the model does not support speech-to-text.