apollo-rate-limits

name	apollo-rate-limits
description	Implement Apollo.io rate limiting and backoff. Use when handling rate limits, implementing retry logic, or optimizing API request throughput. Trigger with phrases like "apollo rate limit", "apollo 429", "apollo throttling", "apollo backoff", "apollo request limits".
allowed-tools	Read, Grep, Bash(curl:*)
version	1.0.0
license	MIT
author	Jeremy Longshore <[e-mail]>

Apollo Rate Limits

Overview

Implement robust rate limiting and backoff strategies for Apollo.io API to maximize throughput while avoiding 429 errors.

Apollo Rate Limits

Endpoint Category	Rate Limit	Window	Burst Limit
People Search	100/min	1 minute	10/sec
Person Enrichment	100/min	1 minute	10/sec
Organization Enrichment	100/min	1 minute	10/sec
Sequences/Campaigns	50/min	1 minute	5/sec
Bulk Operations	10/min	1 minute	2/sec
General API	100/min	1 minute	10/sec

Rate Limit Headers

# Check current rate limit status
curl -I -X POST "https://api.apollo.io/v1/people/search" \
  -H "Content-Type: application/json" \
  -d '{"api_key": "'$APOLLO_API_KEY'", "per_page": 1}'

# Response headers:
# X-RateLimit-Limit: 100
# X-RateLimit-Remaining: 95
# X-RateLimit-Reset: 1640000000
# Retry-After: 60  (only when rate limited)

Implementation: Rate Limiter Class

// src/lib/apollo/rate-limiter.ts
interface RateLimiterConfig {
  maxRequests: number;
  windowMs: number;
  minSpacingMs: number;
}

class RateLimiter {
  private queue: Array<{
    resolve: (value: void) => void;
    reject: (error: Error) => void;
  }> = [];
  private requestTimestamps: number[] = [];
  private lastRequestTime = 0;
  private processing = false;

  constructor(private config: RateLimiterConfig) {}

  async acquire(): Promise<void> {
    return new Promise((resolve, reject) => {
      this.queue.push({ resolve, reject });
      this.processQueue();
    });
  }

  private  () {
     (. || .. === ) ;
    . = ;

     (.. > ) {
      
       now = .();
      . = ..(
         now - ts < ..
      );

      
       (.. >= ..) {
         oldestTs = .[];
         waitTime = .. - (now - oldestTs) + ;
         .(waitTime);
        ;
      }

      
       timeSinceLastRequest = now - .;
       (timeSinceLastRequest < ..) {
         .(.. - timeSinceLastRequest);
      }

      
       item = ..()!;
      ..(.());
      . = .();
      item.();
    }

    . = ;
  }

   (: ): <> {
      ( (resolve, ms));
  }
}


  apolloRateLimiter =  ({
  : , 
  : ,
  : , 
});

Implementation: Exponential Backoff

// src/lib/apollo/backoff.ts
interface BackoffConfig {
  initialDelayMs: number;
  maxDelayMs: number;
  maxRetries: number;
  multiplier: number;
  jitter: boolean;
}

const defaultConfig: BackoffConfig = {
  initialDelayMs: 1000,
  maxDelayMs: 60000,
  maxRetries: 5,
  multiplier: 2,
  jitter: true,
};

export async function withBackoff<T>(
  fn: () => Promise<T>,
  config: Partial<BackoffConfig> = {}
): Promise<T> {
  const cfg = { ...defaultConfig, ...config };
  let lastError: Error;
  let delay = cfg.initialDelayMs;

  for (let attempt = 0; attempt <= cfg.maxRetries; attempt++) {
    try {
      await apolloRateLimiter.acquire();
      return await fn();
    }  (: ) {
      lastError = error;

      
       status = error.?.;
       (status ===  || status ===  || status === ) {
         error; 
      }

       (attempt === cfg.) {
        ;
      }

      
       retryAfter = error.?.?.[];
       (retryAfter) {
        delay = (retryAfter) * ;
      }

      
       jitter = cfg. ? .() *  : ;
       actualDelay = .(delay + jitter, cfg.);

      .();
        ( (r, actualDelay));

      delay *= cfg.;
    }
  }

   lastError!;
}

Implementation: Request Queue

// src/lib/apollo/request-queue.ts
import PQueue from 'p-queue';

// Concurrency-limited queue
export const apolloQueue = new PQueue({
  concurrency: 5, // Max 5 concurrent requests
  interval: 1000, // Per second
  intervalCap: 10, // Max 10 per interval
});

// Usage
async function batchSearchPeople(domains: string[]): Promise<Person[]> {
  const results = await Promise.all(
    domains.map((domain) =>
      apolloQueue.add(() =>
        withBackoff(() => apollo.searchPeople({ q_organization_domains: [domain] }))
      )
    )
  );

  return results.flat().map((r) => r?.people || []).flat();
}

Usage Patterns

Pattern 1: Simple Rate-Limited Request

import { withBackoff } from './backoff';

const people = await withBackoff(() =>
  apollo.searchPeople({
    q_organization_domains: ['stripe.com'],
    per_page: 100,
  })
);

Pattern 2: Batch Processing with Queue

import { apolloQueue } from './request-queue';

async function enrichCompanies(domains: string[]) {
  const results = [];

  for (const domain of domains) {
    const result = await apolloQueue.add(
      () => withBackoff(() => apollo.enrichOrganization(domain)),
      { priority: 1 } // Lower priority
    );
    results.push(result);
  }

  return results;
}

Pattern 3: Priority Queue for Interactive vs Background

// High priority for user-facing requests
async function interactiveSearch(query: string) {
  return apolloQueue.add(
    () => withBackoff(() => apollo.searchPeople({ q_keywords: query })),
    { priority: 0 } // Highest priority
  );
}

// Low priority for background sync
async function backgroundSync(contacts: string[]) {
  return Promise.all(
    contacts.map((id) =>
      apolloQueue.add(
        () => withBackoff(() => apollo.getContact(id)),
        { priority: 10 } // Low priority
      )
    )
  );
}

Monitoring Rate Limit Usage

// src/lib/apollo/rate-monitor.ts
class RateLimitMonitor {
  private requests: Array<{ timestamp: number; remaining: number }> = [];

  recordRequest(remaining: number) {
    this.requests.push({
      timestamp: Date.now(),
      remaining,
    });

    // Keep only last 5 minutes
    const cutoff = Date.now() - 5 * 60 * 1000;
    this.requests = this.requests.filter((r) => r.timestamp > cutoff);
  }

  getStats() {
    const lastMinute = this.requests.filter(
      (r) => r.timestamp > Date.now() - 60000
    );

    return {
      requestsLastMinute: lastMinute.length,
      currentRemaining: lastMinute[lastMinute. - ]?. ?? ,
      : (lastMinute. / ) * ,
      : lastMinute. > ,
    };
  }
}

  rateLimitMonitor =  ();

Output

• Rate limiter class with token bucket algorithm
• Exponential backoff with jitter
• Request queue with concurrency control
• Priority-based request scheduling
• Rate limit monitoring and alerts

Error Handling

Scenario	Strategy
429 response	Use Retry-After header
Burst limit hit	Add minimum spacing
Sustained limit	Queue with concurrency
Network timeout	Exponential backoff

Resources

• Apollo Rate Limits
• p-queue Library
• Exponential Backoff

Next Steps

Proceed to apollo-security-basics for API security best practices.