Vosk Speech Recognition

recognize_vosk(
    audio_data: AudioData,
    verbose: bool = False
) -> str | dict

speech_recognition/
  └── models/
      └── vosk/
          ├── am/           # Acoustic model
          ├── conf/         # Configuration
          ├── graph/        # Language model graph
          └── ...

import speech_recognition as sr

r = sr.Recognizer()

with sr.AudioFile("audio.wav") as source:
    audio = r.record(source)

try:
    text = r.recognize_vosk(audio)
    print(f"Vosk: {text}")
except sr.UnknownValueError:
    print("Vosk could not understand audio")
except sr.RequestError as e:
    print(f"Vosk error; {e}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    print("Speak now...")
    audio = r.listen(source)

print("Transcribing...")
try:
    text = r.recognize_vosk(audio)
    print(f"You said: {text}")
except sr.UnknownValueError:
    print("Could not understand audio")

import speech_recognition as sr
import json

r = sr.Recognizer()

with sr.AudioFile("audio.wav") as source:
    audio = r.record(source)

# Get detailed response
result = r.recognize_vosk(audio, verbose=True)

print(json.dumps(result, indent=2))
print(f"Text: {result['text']}")

# Vosk response typically includes:
# {
#   "text": "transcribed text here"
# }

import speech_recognition as sr

# The model is loaded from speech_recognition/models/vosk/
# Simply replace that directory with a different language model

r = sr.Recognizer()

with sr.AudioFile("german.wav") as source:
    audio = r.record(source)

text = r.recognize_vosk(audio)  # Uses the German model
print(text)

import speech_recognition as sr

r = sr.Recognizer()

with sr.AudioFile("audio.wav") as source:
    audio = r.record(source)

try:
    text = r.recognize_vosk(audio)
    print(f"Transcription: {text}")
    
except sr.UnknownValueError:
    # Speech was unintelligible
    print("Could not understand the audio")
    
except sr.SetupError as e:
    # Model not found or Vosk not installed
    error_msg = str(e).lower()
    if "model not found" in error_msg:
        print("Vosk model not found")
        print("Download with: sprc download vosk")
        print("Or manually from: https://alphacephei.com/vosk/models")
    elif "vosk" in error_msg:
        print("Vosk not installed")
        print("Install with: pip install SpeechRecognition[vosk]")
    else:
        print(f"Setup error: {e}")
        
except Exception as e:
    print(f"Error: {e}")

import speech_recognition as sr

r = sr.Recognizer()

with sr.Microphone() as source:
    print("Adjusting for ambient noise...")
    r.adjust_for_ambient_noise(source, duration=1)
    
    print("Listening... (Press Ctrl+C to stop)")
    
    while True:
        try:
            print("\nSay something:")
            audio = r.listen(source, timeout=5, phrase_time_limit=10)
            
            print("Recognizing...")
            text = r.recognize_vosk(audio)
            
            if text:
                print(f"You said: {text}")
            else:
                print("(silence)")
                
        except sr.WaitTimeoutError:
            print("Listening timed out")
        except sr.UnknownValueError:
            print("Could not understand audio")
        except KeyboardInterrupt:
            print("\nStopping...")
            break

import speech_recognition as sr
import time

r = sr.Recognizer()

with sr.AudioFile("audio.wav") as source:
    audio = r.record(source)

# Vosk (offline)
start = time.time()
vosk_text = r.recognize_vosk(audio)
vosk_time = time.time() - start
print(f"Vosk: {vosk_text} ({vosk_time:.2f}s)")

# Sphinx (offline) - for comparison
start = time.time()
sphinx_text = r.recognize_sphinx(audio)
sphinx_time = time.time() - start
print(f"Sphinx: {sphinx_text} ({sphinx_time:.2f}s)")

# Vosk is typically 2-3x faster than Sphinx with better accuracy