Getting Started With Speech in Azure

# File transcription example (recognize once)
import azure.cognitiveservices.speech as speechsdk

speech_config = speechsdk.SpeechConfig(subscription="your_speech_key", endpoint="https://your-foundry-endpoint/")
audio_config = speechsdk.audio.AudioConfig(filename="record.wav")

recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_config)
result = recognizer.recognize_once_async().get()

if result.reason == speechsdk.ResultReason.RecognizedSpeech:
    print(result.text)
else:
    print(f"Recognition failed: {result.reason}")

# Continuous recognition (microphone) example
import os
import azure.cognitiveservices.speech as speechsdk

endpoint_url = os.environ.get("AZURE_SPEECH_ENDPOINT", "https://your-foundry-endpoint/")
speech_key = os.environ.get("AZURE_SPEECH_KEY", "your_api_key")

speech_config = speechsdk.SpeechConfig(subscription=speech_key, endpoint=endpoint_url)

# Use the default system microphone
audio_config = speechsdk.audio.AudioConfig(use_default_microphone=True)
speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_config)

# Event handlers (handlers receive sender and event arguments)
def recognized_handler(sender, evt):
    # evt.result contains the final recognized text for a segment
    print(f"Recognized: {evt.result.text}")

def recognizing_handler(sender, evt):
    # evt.result contains intermediate partial results
    print(f"Recognizing: {evt.result.text}")

speech_recognizer.recognized.connect(recognized_handler)
speech_recognizer.recognizing.connect(recognizing_handler)

speech_recognizer.start_continuous_recognition()
print("Say something...")

try:
    input("Press Enter to stop...\n")
finally:
    speech_recognizer.stop_continuous_recognition()

# Text-to-speech example (play to default speaker)
import azure.cognitiveservices.speech as speechsdk

speech_config = speechsdk.SpeechConfig(subscription="your_speech_key", endpoint="https://your-foundry-endpoint/")
# Choose a neural voice (example name)
speech_config.speech_synthesis_voice_name = "en-US-AvaNeural"

# No AudioConfig required for default speaker output; provide one for file output instead
speech_synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config)

result = speech_synthesizer.speak_text_async("My name is John").get()

if result.reason == speechsdk.ResultReason.SynthesizingAudioCompleted:
    print("Speech synthesized successfully.")
else:
    print(f"Speech synthesis failed: {result.reason}")