Whisper

# pip install openai-whisper (원본)
# pip install faster-whisper (최적화 버전, 권장)

# === 방법 1: 원본 Whisper (로컬) ===
import whisper

model = whisper.load_model("base")  # tiny, base, small, medium, large-v3

# 음성 파일 전사
result = model.transcribe("audio.mp3", language="ko")
print(result["text"])

# 영어로 번역
result_en = model.transcribe("audio.mp3", task="translate")
print(result_en["text"])

# 타임스탬프 포함
for segment in result["segments"]:
    print(f"[{segment['start']:.2f}s - {segment['end']:.2f}s] {segment['text']}")


# === 방법 2: Faster-Whisper (4배 빠름, 권장) ===
from faster_whisper import WhisperModel

model = WhisperModel("large-v3", device="cuda", compute_type="float16")

segments, info = model.transcribe("audio.mp3", language="ko")

print(f"감지 언어: {info.language} (확률: {info.language_probability:.2f})")

for segment in segments:
    print(f"[{segment.start:.2f}s -> {segment.end:.2f}s] {segment.text}")


# === 방법 3: OpenAI API ===
from openai import OpenAI

client = OpenAI(api_key="your-api-key")

# 전사 (Transcription)
audio_file = open("audio.mp3", "rb")
transcript = client.audio.transcriptions.create(
    model="whisper-1",
    file=audio_file,
    response_format="verbose_json",
    timestamp_granularities=["segment"]
)

print(transcript.text)
for segment in transcript.segments:
    print(f"[{segment.start}s] {segment.text}")

# 번역 (Translation to English)
translation = client.audio.translations.create(
    model="whisper-1",
    file=audio_file
)
print(translation.text)


# === 방법 4: 스트리밍/실시간 전사 (RealtimeSTT) ===
# pip install RealtimeSTT
# from RealtimeSTT import AudioToTextRecorder
#
# recorder = AudioToTextRecorder(model="base")
#
# def process_text(text):
#     print(f"인식된 텍스트: {text}")
#
# recorder.text_callback = process_text
# recorder.start()