• English

      • Subcommand Details

      asr - Speech Recognition

      Recognition Types

      TypeDescriptionUse Case
      onlineReal-time streaming, output as you speakMicrophone, live, low latency
      offlineNon-streaming, processes full audioFile recognition, long audio
      2passDual-stream fusion, real-time + refinementBalance of speed and accuracy

      Basic Usage

      # File recognition (Offline)
manyspeech asr -t offline -f audio.wav

# Microphone recognition (Online)
manyspeech asr -t online -i mic

# 2pass recognition
manyspeech asr -t 2pass -i mic

      Output Formats

      FormatDescription
      textPlain text (Default)
      jsonJSON format with timestamps
      srtSubRip Subtitles
      vttWebVTT Subtitles
      manyspeech asr -t offline --format srt -f audio.wav

      vad - Voice Activity Detection

      Basic Usage

      # Microphone online detection
manyspeech vad -t online -i mic

# Audio file offline detection
manyspeech vad -t offline -f audio.wav

# Specify output format
manyspeech vad -t offline --format wav -f audio.wav

      punc - Punctuation Restoration

      Basic Usage

      # Direct text input
manyspeech punc --text "hello world how are you"

# Read from file
manyspeech punc -f text.txt