ai-rate-queue

Redis-backed requests-per-minute (RPM) limiter for AI/LLM API calls. Works across multiple workers/processes by sharing counters in Redis.

Install

Requires Node.js >=22.

npm install ai-rate-queue

Bring your own Redis client (example with ioredis):

npm install ioredis

Usage

import Redis from "ioredis";
import { createRateLimitQueue } from "ai-rate-queue";

const redis = new Redis(process.env.REDIS_URL!);

const queue = createRateLimitQueue({
  redis,
  requestsPerMinute: 60,
  keyPrefix: "my-app:openai"
});

const result = await queue.enqueue(async () => {
  // call your LLM provider here
  return "ok";
});

Docs

• docs/quickstart.md
• docs/usage.md
• docs/redis.md
• docs/troubleshooting.md