ISpeechToTextTask Interface
Provider-agnostic Speech-to-Text. Implementations configure models on the asset.
Task< string > TranscribeAsync ( byte[] audioBytes,
string language,
CancellationToken ct )
ParametersaudioBytesEncoded audio (e.g., WAV/MP3/OGG).
Providers decide supported formats.
ParameterslanguageOptional BCP-47 / ISO code like "en" (implementations may ignore).