audio_processing.py

import whisper
import tempfile
from pydub import AudioSegment
import os

# Set FFmpeg path manually (if needed)
# AudioSegment.converter = r"C:\ffmpeg\bin\ffmpeg.exe"
AudioSegment.converter = "/usr/bin/ffmpeg"


def transcribe_audio(audio_file):
    model = whisper.load_model("base")

    # Convert to WAV if necessary
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_wav:
        audio = AudioSegment.from_file(audio_file)
        audio.export(temp_wav.name, format="wav")

        result = model.transcribe(temp_wav.name)
        return result["text"]