Skip to content
← Back to News

Anthropic + Cloudflare Distributed Crawling

Cloudflare's edge network now distributes ClaudeBot crawling, improving speed and reducing server load.

TL;DR

Anthropic is moving ClaudeBot crawling to Cloudflare's global edge network (300+ data centers). Expect faster crawl times, more distributed IP ranges, and higher crawl volume for Cloudflare customers.

What the partnership changes for crawling

Previously, ClaudeBot operated from Anthropic's own crawler infrastructure — a centralized system with geographic and capacity limitations. With Cloudflare's edge network:

  • Faster discovery — Content is crawled from edge locations closer to the source
  • Higher volume — Distributed crawling handles burst demand more effectively
  • More IP ranges — Expect ClaudeBot traffic from varied Cloudflare edge IPs
  • Reduced server load — Edge crawling distributes requests more evenly

What to check in your logs

After this transition, monitor your server logs for changes in:

User-agent: ClaudeBot

The ClaudeBot user-agent string should remain consistent; look for traffic from Cloudflare IP ranges.

Increased crawl frequency

Distributed crawling may increase the frequency of requests — check if your rate limits need adjustment.

robots.txt compliance

ClaudeBot should continue respecting robots.txt directives. Verify your rules aren't inadvertently blocking important content.

Action items for website owners

  1. Review robots.txt — Ensure ClaudeBot is allowed to access your most important content. Check for overly broad disallow rules.
  2. Monitor crawl logs — Track ClaudeBot activity to understand crawl patterns and detect any issues.
  3. Check rate limits — If you have rate limiting configured, verify it accommodates increased ClaudeBot traffic without blocking legitimate crawls.
  4. Validate llms.txt — With faster crawling, a well- structured llms.txt becomes more important for content prioritization.

Free GazeRank tool

Check your AI crawler access

GazeRank's robots.txt Checker scans for ClaudeBot, GPTBot, PerplexityBot, and Google-Extended rules — flagging any that might accidentally block your content.