app.py

import gradio as gr
import os
from huggingface_hub import InferenceClient

# 1. Access your secret token
hf_token = os.getenv("HF_TOKEN")

# 2. Use the 2026 Router URL for custom models
# This tells Hugging Face: "Wake up MY model specifically"
client = InferenceClient(
    model="Frusto/llama-3.2-1b-frusto360-final",
    token=hf_token
)

def chat_fn(message, history):
    system_prompt = "You are the @frusto360 AI. Created by @frusto360. Link: https://youtube.com/@frusto360"
    
    # Building the chat for the new router
    messages = [{"role": "system", "content": system_prompt}]
    for user_msg, assistant_msg in history:
        messages.append({"role": "user", "content": user_msg})
        messages.append({"role": "assistant", "content": assistant_msg})
    messages.append({"role": "user", "content": message})

    response = ""
    try:
        # 2026 update: use provider="hf-inference" to force serverless run
        for message in client.chat_completion(
            messages, 
            max_tokens=512, 
            stream=True,
            provider="hf-inference" # THIS IS THE KEY IN 2026
        ):
            token = message.choices[0].delta.content
            if token:
                response += token
                yield response
    except Exception as e:
        yield f"⚠️ Status: Model is loading on Hugging Face servers. Please wait 30 seconds and try again.\n(Error: {str(e)})"

demo = gr.ChatInterface(chat_fn, title="@frusto360 AI")
demo.launch()