xSeek/Docs/Perplexity
Back to documentation

Perplexity user agents, explained.

Perplexity ships two crawlers. Understand what each one does, how to control access, and how to get your content cited in answers.

Perplexity · OfficialUpdated Apr 2026~6 min read2 user agents
TL;DR

Perplexity uses PerplexityBot for indexing and Perplexity-User when the AI assistant browses on behalf of a live user. Both are verifiable via the perplexity.ai domain in the user-agent string.

Overview

Perplexity is an answer engine that crawls the web to provide up-to-date cited responses. Two distinct user agents handle two different jobs — one for the search index, one for live user queries.

Knowing which is which matters: blocking the wrong one cuts you out of Perplexity answers entirely.

How to identify the agents

2 crawlers

Perplexity identifies itself with specific user-agent strings. Verify each through a published IP range.

PB

PerplexityBot

Search index · crawler
Search index
Mozilla/5.0 (compatible; PerplexityBot/1.0; +https://www.perplexity.ai/bot)

Perplexity's general-purpose crawler. Builds and updates the index that powers cited answers. Block this one and you disappear from Perplexity's results.

Verifiableperplexity.ai/bot
PU

Perplexity-User

Live user queries · on-demand
User-triggered
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Perplexity-User/1.0; +https://www.perplexity.ai)

Fires when Perplexity's AI assistant browses the web in response to a live user query. The strongest signal Perplexity is serving your content to a real person.

Verifiableperplexity.ai

How Perplexity accesses your content

3 access patterns, mapped to the agents above.

1

Search index

PerplexityBot crawls content to build and refresh the index behind every answer.
2

Live browsing

Perplexity-User fetches pages in real time when a user asks a current question.
3

Verification

Both identify via perplexity.ai in the UA and can be verified with reverse DNS.

Control Perplexity's access

robots.txt

You control each agent independently. Common configurations below.

Allow all Perplexity accessRecommended

Standard setup if you want to appear in cited answers.
# Allow Perplexity to crawl and cite User-agent: PerplexityBot Allow: / User-agent: Perplexity-User Allow: /

Block Perplexity entirelyRestrictive

Keeps Perplexity out completely — no index, no citations.
# Block every Perplexity crawler User-agent: PerplexityBot Disallow: / User-agent: Perplexity-User Disallow: /

Block sensitive sections only

Keep public content indexable, protect private areas.
# Restrict Perplexity to public content User-agent: PerplexityBot Disallow: /private/ Disallow: /members-only/ Allow: /

Optimize content for Perplexity.

Perplexity rewards content that answers fast and cites well. Five things that work.

Lead with the answer. Perplexity extracts direct answers — the first 2 sentences of a page should resolve the question.
Cite your sources. Outbound citations raise trust signals. Perplexity's model is trained on citation-rich content.
Structure with semantic HTML. Proper headings, lists, and tables so the extractor can isolate the relevant chunk.
Keep it current. Perplexity prefers recent sources for time-sensitive queries. Freshness matters.
Use schema markup. Article, FAQPage, and Dataset structured data help Perplexity verify what your page is about.

Track Perplexity visits with xSeek.

See every Perplexity visit in real time.

Monitor PerplexityBot and Perplexity-User. Track which URLs they hit. See how your content surfaces in Perplexity answers and which queries send traffic back.

Start free →

Frequently asked questions

Two: PerplexityBot/1.0 for web indexing and Perplexity-User/1.0 for live assistant browsing. Both include perplexity.ai in the UA string and can be verified via reverse DNS.
Source: information in this guide is drawn from Perplexity's official bot documentation.
PreviousClaude user agentsNextMistral user agents