Quick Start Guide

from fastapi import FastAPI, Request, HTTPException
from fastapi.middleware.cors import CORSMiddleware
import time
from collections import defaultdict
import asyncio

app = FastAPI()

# Rate limiter storage
request_counts = defaultdict(lambda: {"count": 0, "window_start": time.time()})
RATE_LIMIT = 100  # requests per window
WINDOW_SIZE = 3600  # 1 hour
BURST_LIMIT = 10  # requests per minute

class RateLimiter:
    @staticmethod
    async def check_rate_limit(client_ip: str):
        current_time = time.time()
        client_data = request_counts[client_ip]
        
        # Reset window if expired
        if current_time - client_data["window_start"] > WINDOW_SIZE:
            client_data["count"] = 0
            client_data["window_start"] = current_time
        
        # Check burst limit (last minute)
        recent_requests = client_data.get("recent", [])
        recent_requests = [t for t in recent_requests if current_time - t < 60]
        
        if len(recent_requests) >= BURST_LIMIT:
            raise HTTPException(status_code=429, detail="Burst limit exceeded")
        
        # Check window limit
        if client_data["count"] >= RATE_LIMIT:
            raise HTTPException(status_code=429, detail="Rate limit exceeded")
        
        # Update counts
        client_data["count"] += 1
        recent_requests.append(current_time)
        client_data["recent"] = recent_requests

@app.middleware("http")
async def rate_limit_middleware(request: Request, call_next):
    client_ip = request.client.host
    
    # Skip rate limiting for health checks
    if request.url.path == "/health":
        return await call_next(request)
    
    await RateLimiter.check_rate_limit(client_ip)
    response = await call_next(request)
    return response

@app.post("/api/predict")
async def predict(request: Request):
    # Your AI model endpoint
    return {"status": "success"}