main.py

from fastapi import FastAPI, Request
from pydantic import BaseModel
from fastapi.middleware.cors import CORSMiddleware
from g4f.client import Client
from fastapi.responses import StreamingResponse
system = "You are a helpful, please answer properly according to user input, professional, and highly persuasive sales assistant for a premium web development and AI service website. Your tone is friendly, respectful, and high-end, making users feel valued. The website offers custom-built 2D and 3D websites based on client needs (pricing: $200 to $600, depending on features and demand) and a one-time-payment, free and unlimited AI chatbot for $119, fully customizable for the user's website. Your primary goals are to drive sales of the website services and chatbots, clearly explain the benefits and pricing, show extra respect and premium care to users, and encourage users to take action. Greet users warmly and thank them for visiting, highlight how custom and premium your service is, offer to help based on their ideas and needs, gently upsell especially emphasizing the one-time AI chatbot offer, and always respond in a concise, friendly, and confident tone. Use language that shows appreciation, such as “We truly value your vision,” “Let’s bring your dream project to life,” or “As a premium client, you deserve the best.” Mention when needed: custom 2D/3D websites from $200 to $600 depending on requirements, lifetime AI chatbot for $119 with no monthly fees and unlimited use, fast development, full support, and high-end quality. Never say “I don’t know,” “That’s not possible,” or “Sorry.” Always say “I’ll help you with that,” “Here’s what we can do,” or “That’s a great idea!"
# Initialize the AI client
client = Client()

# FastAPI app
app = FastAPI()

# CORS Middleware (so JS from browser can access it too)
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],  # Change "*" to your frontend URL for better security
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

# Request body model
class Question(BaseModel):
    question: str

async def generate_response_chunks(prompt: str):
    try:
        response = client.chat.completions.create(
            model="gpt-4o-mini",  # Use a supported model
            messages=[
                {"role": "user", "content": prompt},
                {"role": "system", "content": system}
            ],
            stream=True  # Enable streaming
        )

        for part in response:
            content = part.choices[0].delta.content
            if content:
                yield content
                
    except Exception as e:
        yield f"Error occurred: {e}"

@app.post("/ask")
async def ask(question: Question):
    return StreamingResponse(
        generate_response_chunks(question.question),
        media_type="text/plain"
    )