import os from fastapi import FastAPI, Request from pydantic import BaseModel from openai import OpenAI app = FastAPI() client = OpenAI( base_url="https://router.huggingface.co/v1", api_key=os.environ["HF_TOKEN"], ) class Query(BaseModel): question: str @app.get("/") async def root(): return {"message": "API está rodando!"} @app.post("/ask") async def ask_model(query: Query): completion = client.chat.completions.create( model="Qwen/Qwen3-4B-Thinking-2507:nscale", messages=[ {"role": "user", "content": query.question} ], ) answer = completion.choices[0].message.content return {"answer": answer}