UnsourcedSee who crawls your site →

AI SearchPerplexity

Perplexity-User: IP ranges, verification & how to handle it

Last verified: 2026-07-16 · maintained by Unsourced

Perplexity-User is the on-demand retrieval that runs the instant a person's question needs a page the index hasn't already covered. It plays no cataloguing role — it reads, answers, and moves on.

What Perplexity-User does, and how it differs

Perplexity-User keeps no library and follows no schedule: it activates only when someone's specific prompt points at a URL the standing index doesn't already hold, fetches that page in the moment, and folds it straight into the reply they're waiting on. Each hit therefore maps to a real person mid-task rather than a maintenance sweep. Blocking it doesn't trim a dataset — it pulls pages out of live answers at the exact instant a human asked to see them.

How to verify Perplexity-User

Perplexity doesn't expose a reverse-DNS hostname for Perplexity-User, so verification rests on its 12 published IP ranges: the request is genuine only when its source address sits inside that set. A Perplexity-User user-agent arriving from any other IP is spoofed — on its own the header proves nothing.

Should you allow or block Perplexity-User?

Recommended: bait. Blocking it removes nothing from a dataset — it strips your page out of the specific live answer a person is waiting on, so the cost is paid in real time rather than in future training.

If you do choose to act in robots.txt (which crawlers honour but don't enforce):

# Perplexity-User: recommended to ALLOW — blocking can cost you AI visibility
User-agent: Perplexity-User
Disallow:

Official sources

Common questions about Perplexity-User

When does Perplexity-User fetch my page?

Only when someone's specific question points at a URL Perplexity's standing index doesn't already hold. It reads the page in the moment, folds it into the reply, and moves on — there's no cataloguing.

Does Perplexity-User keep an index of my site?

No. It holds no library and follows no schedule — each hit is a live, one-off retrieval tied to a real person's prompt, unlike PerplexityBot which builds the standing index.

Should I block Perplexity-User?

Blocking it doesn't trim a dataset — it pulls your page out of the specific live answer a person is waiting on, so the cost lands in real time. For most sites that visibility is worth keeping.

Related crawlers

Stop trusting the Perplexity-User user-agent.

Unsourced tells you which requests claiming to be Perplexity-User are verified, which are impostors, and which AI assistants actually link to your pages.

Check your site free →

10-day free trial · no card required · cancel anytime