Operator
Meta
Meta runs 16 bots across 5 agent types. Each one is a separate user-agent so you can allow or block them independently.
CrowdTangle is operated by Meta. Maps link graphs, ranking signals, and audit data for SEO tooling. Typically run by SaaS platforms on behalf of marketers monitoring their own or competitor sites.
facebook is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
FacebookBot is operated by Meta. Crawls broadly across the public web to gather content for training new model versions. Site owners can opt out via robots.txt, though enforcement varies by operator.
FacebookCatalog is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
FacebookExternalHit is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
FacebookExternalUserAgent is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
Facebot is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
Instagram is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
Llama-Stack-Bot is operated by Meta. Fetches a page in real time when an LLM is mid-answer to a user question. Traffic spikes correlate with question volume rather than crawl schedules, so it shows up unpredictably.
meta-externalads is operated by Meta. Crawls broadly across the public web to gather content for training new model versions. Site owners can opt out via robots.txt, though enforcement varies by operator.
Meta-ExternalAgent is operated by Meta. Crawls broadly across the public web to gather content for training new model versions. Site owners can opt out via robots.txt, though enforcement varies by operator.
Meta-ExternalFetcher is operated by Meta. Fetches a page in real time when an LLM is mid-answer to a user question. Traffic spikes correlate with question volume rather than crawl schedules, so it shows up unpredictably.
meta-webindexer is operated by Meta. Indexes pages for AI-powered search products that cite and link back to sources. Allowing it is typically how a site becomes citable inside answer engines and LLM search.
Threadbot is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
Threadsbot is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
WhatsApp is operated by Meta. Single-page fetch triggered by a user action, link previews, social unfurlers, RSS readers. Volume tracks shares and clicks rather than crawl schedules.
Live data · via Cloudflare Radar
See how Meta's bots show up in real internet traffic
Cloudflare Radar publishes a continuously-updating view of AI bot activity across the websites Cloudflare protects. You can see which user-agents drive the most requests, how that share shifts week to week, and which industries each bot crawls hardest. The data refreshes on Cloudflare's side, so the link below is always current.
Open Cloudflare Radar AI InsightsWhat you'll find there
- ·User-agent share for every tracked AI bot, including Meta's
- ·Robots.txt and crawl-purpose breakdowns per operator
- ·Industry and content-type targets for each crawler
- ·Time-series trends going back several months