The native / built-in API SpeechAnalyzer (introduced last generation on iOS and MacOS) works quite well.

Are folks finding it lacking / needing alternatives?