---
title: "Perplexity-User vs PerplexityBot: Which One Hit Your Site"
description: "Perplexity-User is Perplexity's user-triggered fetcher; here is what it does, how it treats robots.txt, and how it differs from PerplexityBot."
url: https://citaeotion.ai/perplexity-user-vs-perplexitybot/
date: 2026-10-07
modified: 2026-09-30
author: "citAEOtion"
image: https://citaeotion.ai/wp-content/uploads/2026/09/perplexity-user-vs-perplexitybot-featured.webp
categories: ["AI Traffic Intelligence", "Answer Engine Optimization"]
type: post
lang: en
---

# Perplexity-User vs PerplexityBot: Which One Hit Your Site

Perplexity-User is the agent Perplexity uses to visit a web page when a person asks Perplexity a question and it needs that page to answer. It is a user-triggered fetch, not a crawler, and Perplexity says it is not used to collect content for training AI foundation models. Because a person requested the fetch, Perplexity says it generally ignores robots.txt rules.

**Quick answer:** Perplexity runs two documented agents. PerplexityBot crawls the web to surface and link sites in Perplexity's search results and follows robots.txt. Perplexity-User fetches a specific page in real time when someone asks a question, may include a link to that page in the answer, and generally ignores robots.txt because the fetch was user-requested. Neither is used to train foundation models, according to Perplexity. In your logs, PerplexityBot looks like a crawl across many URLs, while Perplexity-User shows up as scattered hits on the pages real questions are pulling.

## What Perplexity-User does

Perplexity's [bots documentation](https://docs.perplexity.ai/guides/bots) says Perplexity-User "supports user actions within Perplexity." When someone asks a question, it "might visit a web page to help provide an accurate answer and include a link to the page in its response."

That makes it the agent closest to an actual citation. A Perplexity-User request means a real question led Perplexity to your page. It does not prove the page was cited, but it tells you your content was pulled into the answering process for a live query. If you want to understand how those answers pick sources, our guide on [how to get cited by Perplexity](/how-to-get-cited-by-perplexity/) goes through the content side.

The documentation is also clear on what Perplexity-User is not. It states that the agent "is not used for web crawling or to collect content for training AI foundation models."

### The robots.txt exception

This is the part most site owners miss. Perplexity's documentation says: "Since a user requested the fetch, this fetcher generally ignores robots.txt rules." So a Disallow line aimed at Perplexity-User will generally not stop it.

That puts Perplexity-User in a different category from a crawler. The reasoning is that a person asked for the page, much like a person clicking a link. Whether you agree with that reasoning or not, it means robots.txt is not the control to reach for if you want to limit these fetches. Server-level or firewall rules, matched against the IP ranges Perplexity publishes, are the usual route. Our guide to [AI crawler access control](/ai-crawler-access-control/) covers the options.

### Full user agent string

Perplexity documents the full string as:

```
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Perplexity-User/1.0; +https://perplexity.ai/perplexity-user)
```

Perplexity also publishes a separate list of IP ranges for Perplexity-User, linked from the same documentation page. Use it to confirm that a request carrying this name really comes from Perplexity.

## How PerplexityBot differs

PerplexityBot is the crawler. Perplexity's documentation says it is "designed to surface and link websites in search results on Perplexity" and is not used to crawl content for AI foundation models. Perplexity recommends allowing PerplexityBot in robots.txt and permitting requests from its published IP ranges if you want to appear in its search results. For the full picture of that crawler, see our guide to [PerplexityBot](/perplexitybot-explained/).

Its documented user agent string is:

```
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot)
```

## Perplexity-User vs PerplexityBot at a glance

| Signal | Perplexity-User | PerplexityBot |
| --- | --- | --- |
| Documented purpose | Fetches a page when a user asks a question, may link it in the answer | Surfaces and links websites in Perplexity search results |
| Is it a crawler? | No, per Perplexity | Yes |
| Used to train foundation models | No, per Perplexity | No, per Perplexity |
| robots.txt | Generally ignored, because the fetch is user-requested | Perplexity recommends allowing it in robots.txt |
| What triggers it | A live question from someone using Perplexity | Indexing work for search coverage |
| How to verify | Match the IP against Perplexity's published Perplexity-User ranges | Match the IP against Perplexity's published PerplexityBot ranges |

Read the table as two questions. PerplexityBot answers "is my site in Perplexity's index?" Perplexity-User answers "is a real question pulling this specific page right now?"

## How to spot Perplexity-User in your server logs

Start with the user agent field, then check the details:

- **Find the token.** Search for "Perplexity-User" in the user agent string. It is distinct from "PerplexityBot."

- **Look at the pattern.** A crawler tends to work through many URLs. User-triggered fetches tend to land on individual pages that match questions people are asking, and they can come and go with interest in a topic.

- **Note which pages.** The pages Perplexity-User requests most are the pages real questions are pulling. Those are strong candidates to refresh and expand.

- **Check the status code.** A Perplexity-User request that gets a 404, a 403, or a slow redirect chain is a question your site failed to answer. Our guide to [HTTP status codes for AI crawlers](/http-status-codes-ai-crawlers/) explains each response.

- **Verify the IP.** Compare the requesting address against Perplexity's published Perplexity-User ranges.

That last step matters. Any client can copy a user agent string, so a log line reading Perplexity-User is a claim, not proof. Checking against the operator's published ranges is how you [separate a real agent from a spoofed one](/verify-ai-crawler-vs-spoofed-bot/).

## Why a Perplexity-User hit is not the same as a citation

A Perplexity-User request tells you your page was fetched while a question was being answered. Whether it made it into the final answer, and whether the link was shown, is a separate outcome your logs cannot confirm on their own.

Still, it is one of the most useful signals you can get. It is observed behavior, not a simulation. Prompt-based tools model what users might ask and infer visibility from sample answers. Your logs record what actually happened on your domain: which page, when, and whether the server said yes.

## Where Perplexity-User fits among AI agents

Perplexity is not alone in splitting its traffic this way. Several AI companies now run separate agents for training, for search indexing, and for fetches triggered by a user's question. Each category deserves its own decision. Our guide to [AI crawler categories](/ai-crawler-categories/) lays out the three groups and how they tend to treat robots.txt.

User-triggered fetchers are the category most tied to live answers. If you block them, you may be removing your pages from answers to questions people are asking right now. If you allow them, you want to know which pages they request and whether those pages load cleanly.

## Practical steps

1. Decide whether you want to appear in Perplexity's search results. If yes, allow PerplexityBot in robots.txt and permit its published IP ranges, as Perplexity recommends.
2. Treat Perplexity-User as a separate decision. Remember that robots.txt generally does not control it, so any restriction has to happen at the server or firewall level.
3. Verify both agents against their separate published IP lists before acting on any hit.
4. Track Perplexity-User requests per page to see which content real questions pull most often.
5. Fix any 4xx or 5xx responses on those requests first. They are lost answers.
6. Recheck Perplexity's bots documentation periodically. Agent names, behavior, and IP ranges can change, and your policy should follow the source.

To see Perplexity-User and PerplexityBot requests side by side, per page and per status code, [citAEOtion is a $99 Lifetime Deal](/pricing/), or you can [book a 20-minute walkthrough](/demo/) first.

## Frequently Asked Questions

### Does Perplexity-User follow robots.txt?

Generally, no. Perplexity's documentation says that since a user requested the fetch, Perplexity-User generally ignores robots.txt rules. If you need to limit it, use server or firewall rules matched to Perplexity's published Perplexity-User IP ranges.

### Does Perplexity-User train AI models on my content?

According to Perplexity, no. Its documentation states that Perplexity-User is not used for web crawling or to collect content for training AI foundation models. It fetches pages to help answer a user's question.

### Why do I see Perplexity-User in my logs but no PerplexityBot?

The two agents do different jobs. Perplexity-User only appears when someone asks a question that leads to your page. If PerplexityBot is blocked in robots.txt, or has not reached your site yet, the crawl side stays quiet while user-triggered fetches can still arrive, since those generally ignore robots.txt.

### Does a Perplexity-User visit mean my page was cited?

Not necessarily. Perplexity says the agent may include a link to the page in its response, but a fetch alone does not confirm a citation. Treat each hit as evidence your page was considered for a real question, and use [per-page crawler tracking](/ai-crawler-tracking/) to see which pages get pulled most.

### How do I tell a real Perplexity-User request from a fake one?

Do not trust the user agent string alone, because anyone can copy it. Compare the requesting IP address with the Perplexity-User IP ranges linked from Perplexity's bots documentation. A request with the right name but an unlisted IP should be treated as an unknown bot.