LLM_Handler.py

import openai
import os
from dotenv import load_dotenv

# Load OpenAI API key from .env file
load_dotenv()
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

# Check if the API key is loaded correctly
if not OPENAI_API_KEY:
    raise ValueError("❌ OpenAI API Key is missing! Please set it in your .env file.")

# Initialize OpenAI client
client = openai.OpenAI(api_key=OPENAI_API_KEY)

def process_query(audio=None, text=None):
    """
    Handles both audio and text input.
    1. If audio is provided, transcribes it using Whisper.
    2. If text is provided, processes it with GPT-4o.
    Returns: (transcript_text, audio_output_url)
    """

    transcript = None
    if audio:
        try:
            with open(audio, "rb") as audio_file:
                response = client.audio.transcriptions.create(
                    model="whisper-1",
                    file=audio_file
                )
            transcript = response.text
        except Exception as e:
            transcript = f"Error in transcription: {str(e)}"

    # Use text input or transcribed text for GPT-4o response
    query_text = text if text else transcript

    if not query_text:
        return "No valid input provided!", None

    # Get response from GPT-4o
    try:
        gpt_response = client.chat.completions.create(
            model="gpt-4o",
            messages=[{"role": "system", "content": "You are an AI assistant."},
                      {"role": "user", "content": query_text}]
        )
        ai_text_response = gpt_response.choices[0].message.content
    except Exception as e:
        ai_text_response = f"Error in GPT response: {str(e)}"

    # Generate AI voice response
    try:
        audio_response = client.audio.speech.create(
            model="tts-1",
            voice="alloy",
            input=ai_text_response
        )
        audio_path = "response_audio.mp3"
        with open(audio_path, "wb") as audio_file:
            audio_file.write(audio_response.content)
    except Exception as e:
        audio_path = None
        ai_text_response += f"\nError in generating audio: {str(e)}"

    return ai_text_response, audio_path