Glm Pricing

import asyncio
import random
import time
from typing import Callable, Any

class AsyncRateLimiter:
    """Production-ready async rate limiter with jittered exponential backoff."""
    def __init__(self, max_requests: int = 60, time_window: float = 60.0):
        self.max_requests = max_requests
        self.time_window = time_window
        self.tokens = max_requests
        self.updated_at = time.monotonic()
        self._lock = asyncio.Lock()

    async def acquire(self):
        async with self._lock:
            now = time.monotonic()
            elapsed = now - self.updated_at
            self.tokens = min(self.max_requests, self.tokens + elapsed * (self.max_requests / self.time_window))
            self.updated_at = now
            if self.tokens < 1.0:
                sleep_needed = (1.0 - self.tokens) * (self.time_window / self.max_requests)
                await asyncio.sleep(sleep_needed + random.uniform(0.01, 0.05))
            self.tokens -= 1.0

async def execute_task(limiter: AsyncRateLimiter, task_id: int):
    await limiter.acquire()
    print(f"[Done] Task {task_id} completed successfully")