Skip to content
AI bots / Perplexity

Perplexity User: a live read for one question

Perplexity-User is the agent Perplexity sends when a user asks a question and the answer needs your page. Perplexity may then include a link to that page in its response. It is not a crawler, and Perplexity says it does not collect training content.

One detail sets it apart from its sibling crawler. Perplexity states that this fetcher generally ignores robots.txt, because a person asked for the fetch.

The Pagereckon team Updated
Perplexity user at a glance, from Pagereckon
User-agent string
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Perplexity-User/1.0; +https://perplexity.ai/perplexity-user)
Operator
Perplexity
Type
Live fetch for a user
robots.txt token
Perplexity-User
Obeys robots.txt
No. Perplexity says the fetcher generally ignores robots.txt, since a user requested the fetch
Official docs
docs.perplexity.ai

Why robots.txt does not stop a Perplexity user fetch

The logic follows from how the agent works. A person asks a question, and Perplexity opens a page on their behalf, much like a browser would. So Perplexity treats it as a user action, not as automated crawling.

That is a clear, published position, and it differs from Anthropic's live fetcher, which follows robots.txt. If you need Perplexity to stay away from a page, you must refuse the request at the server.

.htaccess (Apache, LiteSpeed): answers 403
<IfModule mod_rewrite.c>
RewriteEngine On
RewriteCond %{HTTP_USER_AGENT} Perplexity-User [NC]
RewriteRule .* - [F,L]
</IfModule>
Nginx server block: answers 403
if ($http_user_agent ~* "Perplexity-User") {
    return 403;
}

Pagereckon Pro's one-click blocking covers bots that crawl hard and give nothing back, such as Bytespider. A live fetcher that can send readers is not on that list, so you add the rule above by hand if you want it.

What you lose when you block it

When a Perplexity user asks about your topic, the answer can no longer read your page or link to it. Instead, it will lean on other sources. Since every Perplexity answer lists its sources, that is a lost citation and often a lost visit.

Also, the block does not change your place in Perplexity's index. That depends on PerplexityBot, which follows robots.txt and has its own token.

Verify a Perplexity user request

Perplexity publishes the IP ranges for this agent at perplexity.com/perplexity-user.json. Take the IP from your log line and check it against that list. Perplexity suggests combining the user agent check with the IP check, and refreshing the list regularly.

When the address does not match, the request did not come from Perplexity. Some scrapers use AI agent names because they expect site owners to let those through.

Perplexity user reads in Pagereckon

Pagereckon files the agent under live reads. Every live read stands for a person who asked Perplexity something your page could answer. Training crawls and AI search crawls sit in separate groups, so this number stays clean.

For example, a guide that gets ten live reads a day but no visits from perplexity.ai is being read and quoted, yet rarely clicked. That is useful to know before you edit it. See the split in the AI crawler checker, and the visits in AI referral traffic.

Perplexity user agent questions

Does Perplexity-User respect robots.txt?

Generally no. Perplexity says the fetcher generally ignores robots.txt, because a user requested the fetch.

Is a Perplexity user fetch a visitor?

No. It is the assistant reading your page. A visitor arrives only when the person clicks the link in the answer.

Does Perplexity use these fetches for training?

Perplexity says no. The agent is not used for crawling or for collecting content to train AI foundation models.

How do I stop it on one page only?

Add a server rule that matches both the user agent and that page's path, then answer 403.

Sources and references

Free plugin, optional Pro

See your own AI traffic in ten minutes

Install the free plugin, point it at your access log and read today's humans, bots and AI visits. Then add Pro when you want Search Console and the before and after reports.