GeoPromptTracker

AI crawler comparison

Meta-ExternalAgent vs Meta-ExternalFetcher: what's the difference?

Meta runs one crawler for training data and another for user-triggered link fetches.

Short answer

They are not interchangeable. Meta-ExternalAgent handles AI model training, Meta-ExternalFetcher handles AI search and live answers. Blocking one has no effect on the other, so a robots.txt rule aimed at the wrong agent produces exactly the outcome you were trying to avoid.

Side by side

 Meta-ExternalAgentMeta-ExternalFetcher
OperatorMetaMeta
PurposeAI model trainingAI search & live answers
robots.txtRespects robots.txtIgnores robots.txt
Published IP rangesNone publishedNone published
Official docsYesYes
Blocking costs youNo traffic — training onlyAI-answer citations

What Meta-ExternalAgent does

Meta-ExternalAgent is Meta's crawler for collecting training data for its Llama models and Meta AI products. It is distinct from facebookexternalhit (which fetches link previews), so you can block AI training without breaking how your links render on Facebook and Instagram.

Full profile, with every robots.txt, nginx, Apache and Cloudflare rule: Meta-ExternalAgent.

What Meta-ExternalFetcher does

Meta-ExternalFetcher retrieves individual URLs that Meta AI products need for a user's request — distinct from Meta-ExternalAgent, which crawls for model training. Because its fetches are user-initiated, Meta says this agent may bypass robots.txt rules, making it one of the openly documented cases of a user-fetch agent that doesn't honor blocks.

Full profile, with every robots.txt, nginx, Apache and Cloudflare rule: Meta-ExternalFetcher.

Block both

robots.txt
User-agent: Meta-ExternalAgent
Disallow: /

User-agent: Meta-ExternalFetcher
Disallow: /

Block Meta-ExternalAgent, allow Meta-ExternalFetcher

The split most sites want when the two agents do different jobs: opt out of one without giving up the other.

robots.txt
User-agent: Meta-ExternalAgent
Disallow: /

User-agent: Meta-ExternalFetcher
Allow: /

Check your own site

Frequently asked questions

What is the difference between Meta-ExternalAgent and Meta-ExternalFetcher?

Meta-ExternalAgent is Meta's AI model training agent — Crawls content to train Meta's AI models. Meta-ExternalFetcher is Meta's AI search and live answers agent — Fetches specific links for Meta AI when users share or request them. They are separate user agents with separate robots.txt rules.

Does blocking Meta-ExternalAgent also block Meta-ExternalFetcher?

No. robots.txt matches on the user-agent token, so a group naming Meta-ExternalAgent applies only to Meta-ExternalAgent. Meta-ExternalFetcher reads the group that names it, or the wildcard group if none does. To stop both you need a rule for each — or a wildcard rule, which would also affect every other crawler that reads it.

Should I block Meta-ExternalAgent, Meta-ExternalFetcher, or both?

It depends which outcome you want. Blocking Meta-ExternalAgent opts you out of Meta's model training and costs you no traffic today, because training crawlers send no visitors. Blocking Meta-ExternalFetcher removes you from the answers Meta's assistant shows its users. The common choice is to block training agents and allow search agents, so your content stays citable without feeding model training.

Do Meta-ExternalAgent and Meta-ExternalFetcher both respect robots.txt?

Meta-ExternalAgent: Meta documents the agent and states it respects robots.txt rules. Meta-ExternalFetcher: Meta's own documentation states Meta-ExternalFetcher may bypass robots.txt because fetches are performed on behalf of a user.

Can I tell Meta-ExternalAgent and Meta-ExternalFetcher apart in my server logs?

Yes — they send different user-agent strings, so grep for each token separately: grep -i "Meta-ExternalAgent" and grep -i "Meta-ExternalFetcher". Be aware that a user-agent header is self-declared and anything can send either string. Neither operator publishes IP ranges for these agents, so log entries claiming them cannot be verified.

Other comparisons

Part of our directory of every known AI crawler. Last verified: 2026-08-14.