What is Meta-WebIndexer? Meta's AI crawler
Meta-WebIndexer is the crawler Meta says it uses to improve the quality of Meta AI search results.
129 of the 1662 major websites we measure block Meta-WebIndexer in robots.txt (8%), making it number 27 of the 36 AI crawlers we track for how often it is blocked. 116 of those sites name it directly in their robots.txt; the rest are covered by a wildcard rule.
Key facts
| Fact | Detail |
|---|---|
| Operator | Meta |
| Used for | Meta AI search |
| Type | Search index. Decides whether a site can appear in that product's AI search answers. |
| robots.txt token | Meta-WebIndexer |
| Operator's documentation | Read it here |
What blocking Meta-WebIndexer does
Meta says allowing it helps Meta cite and link to a site's content in Meta AI responses, and that robots.txt changes can take up to 24 hours to apply.
How to allow or block Meta-WebIndexer
Add one of these groups to the robots.txt file at the root of your domain. It is a stated preference that well-behaved crawlers read, not an enforcement mechanism.
User-agent: Meta-WebIndexer Disallow: /
User-agent: Meta-WebIndexer Allow: /
To see whether your own site lets Meta-WebIndexer in, and which rule decides it, run the free AI crawler checker.
Meta-WebIndexer user agent string
Meta's page doesn't list a full string, so look for Meta-WebIndexer in the user agent field of your logs.
How to verify a request is really Meta-WebIndexer
Anyone can put a crawler's name in a user agent string, so the name alone proves nothing. We did not find a published IP list or reverse DNS rule from Meta.
How to whitelist Meta-WebIndexer in a firewall or CDN
robots.txt only states a preference. A firewall, CDN or bot protection service can still refuse the request, and the crawler then gets an error even though robots.txt allows it.
- Check that no group in your robots.txt blocks it, using the rule above.
- Add an allow rule in your firewall or CDN that matches the name in the user agent. Keep the rule narrow, because anyone can copy a name.
- Check your server logs for a 200 response from Meta-WebIndexer afterwards.
On Cloudflare, also check AI Crawler and Agent access policies, which can block search, agent and training traffic across a site.
Why Meta-WebIndexer gets a 403 or 405 error
A 403 means the server refused the request, which usually points to a firewall or bot rule. A 405 means the server refused the request's method, so check any rule that only permits certain methods. A crawler that obeys robots.txt won't request a page it is told to avoid, so an error in your logs points to your server's configuration. To stop it deliberately at the server, this nginx rule returns a 403:
if ($http_user_agent ~* "Meta-WebIndexer") {
return 403;
}Which sites recently changed their rules for Meta-WebIndexer?
None that lasted since 17 August 2026. See the full change log.
Who blocks Meta-WebIndexer, by sector
| Sector | Sites measured | Blocking | Share |
|---|---|---|---|
| News publishers | 182 | 43 | 24% |
| Technology media | 126 | 27 | 21% |
| Social and media platforms | 115 | 23 | 20% |
| Health publishers | 88 | 8 | 9% |
| Reference and research | 114 | 7 | 6% |
| E-commerce and retail | 92 | 5 | 5% |
| Finance and business media | 94 | 5 | 5% |
| AI companies | 128 | 3 | 2% |
| Government and institutions | 98 | 2 | 2% |
| Travel | 86 | 2 | 2% |
| Developer references | 159 | 1 | 1% |
| Hosting and infrastructure | 98 | 1 | 1% |
| SaaS platforms | 196 | 2 | 1% |
| SEO and marketing tools | 86 | 0 | 0% |
Sites blocking Meta-WebIndexer
abc.es, academia.edu, afterpay.com, al.com, allrecipes.com, amazon.ca, amazon.co.uk, amazon.com, amazon.de, androidauthority.com, androidcentral.com, architecturaldigest.com, arstechnica.com, azcentral.com, babycenter.com, barrons.com, belfasttelegraph.co.uk, bonappetit.com, cleveland.com, cntraveler.com, computerbase.de, congress.gov, consilium.europa.eu, dailymail.co.uk, devclass.com, disneyplus.com, distill.pub, dn.se, eltiempo.com, eluniversal.com.mx, epicurious.com, eurogamer.net, everydayhealth.com, ew.com, extremetech.com, facebook.com, faz.net, flickr.com, fortinet.com, ft.com, gamesindustry.biz, globeandmail.com, gq.com, hackernoon.com, hindustantimes.com, hs.fi, ign.com, imdb.com, imore.com, independent.ie, instacart.com, instagram.com, investopedia.com, itpro.com, kiplinger.com, laptopmag.com, lemonde.fr, lesechos.fr, lifehacker.com, linguee.com, and 69 more in the full dataset.
Sites that allow Meta-WebIndexer by name
autotrader.co.uk, beehiiv.com, deezer.com, digitaltrends.com, enterprise.com, expedia.com, fastcompany.com, gainsight.com, getyourguide.com, golem.de, heygen.com, inc.com, khaleejtimes.com, orbitz.com, pcmag.com, premiumtimesng.com, reuters.com, sbs.com.au, snapchat.com, spectrum.ieee.org, vrbo.com.
Other Meta crawlers
Meta-ExternalAgent (model training), Meta-ExternalFetcher (on-demand fetch), facebookexternalhit (link preview), Meta-ExternalAds (advertising products).
See all 36 AI crawlers, the rule change log or the full study. Being reachable is only the first step: how AI search engines choose what to cite covers the rest.
Cite this data
You are welcome to quote or republish these figures. Please credit IndexHalo with a link to this page (CC BY 4.0). All our datasets are on the data page.
IndexHalo. "Meta-WebIndexer block rate across major websites." Measured 12 November 2026. https://www.indexhalo.com/ai-crawlers/meta-webindexer/
<a href="https://www.indexhalo.com/ai-crawlers/meta-webindexer/">Meta-WebIndexer block rate across major websites, IndexHalo</a>