Quick answer
Perplexity is an AI-powered search tool: you ask a question and it replies with an answer that links to numbered web sources. As of October 7, 2026, its own pages describe a free Standard tier, a Pro price printed as "$17 /month or equivalent, when billed annually", an AI-training toggle and how PerplexityBot treats robots.txt.
Key takeaways
- Perplexity answers with numbered source links, and its own help page says it encourages you to "double-check sources".
- Perplexity's pages read on October 7, 2026 name a free Standard tier and print a Pro price, and a footnote under the plan cards reads "For personal use only and subject to our policies." Its plans help article and its Enterprise pricing page print Enterprise Pro seat prices with different billing wording.
- On Free, Pro and Max, an AI data retention toggle controls use of your data for training, and Perplexity says opt-outs apply only to data collected after you opt out.
- Being cited has its own rules: Perplexity's documentation names two user agents, PerplexityBot and Perplexity-User, and says that, "Since a user requested the fetch, this fetcher generally ignores robots.txt rules." for Perplexity-User.
- This site's suggestion: check your robots.txt and firewall for those two names.
What is Perplexity AI?
Perplexity is an AI-powered search tool: you type a question and it searches the web and replies with a written answer that cites its sources.
The wording is this site's plain summary; Perplexity's help center calls it an "AI-powered search engine". Perplexity's own description (its wording, not this site's) says: "Simply ask any question, and it searches the web to deliver accessible, conversational answers backed by verifiable sources." (Help center, read October 7, 2026.)
On how it differs from a traditional search engine, this is Perplexity's own marketing wording, not this site's: "Unlike traditional search engines, which make you sift through a list of links, Perplexity delivers the insights you’re looking for in one place." (Answer-engine help article, last modified September 3, 2026.)
For an owner, the uses the help center names are "seeking quick answers to everyday questions" and "conducting in-depth research for professional work".
Disclosure: drafted with Claude, made by Anthropic, a competitor of Perplexity's in AI assistants.
How do you check a source that Perplexity cites?
Open the numbered link next to the sentence you plan to reuse, then find that claim on the original page. Perplexity's help center says: "Each answer includes numbered citations linking to the original sources, allowing you to easily verify the information or explore further." Another help page adds: "While we aim for accuracy, we encourage you to double-check sources for added confidence."
This site's two-click check (a suggestion, not Perplexity's instruction):
- Click the number beside the sentence to open the linked page.
- Find the supporting passage and note who published it and when. If you cannot find it, treat the sentence as unconfirmed.
What do Perplexity's own pages say about mistakes?
They say to double-check, and the Terms of Service say the output may be wrong. None of the pages read gives an accuracy rate.
| Where | What it says, literally |
|---|---|
| Help center, answer-engine article | "While we aim for accuracy, we encourage you to double-check sources for added confidence." |
| Terms of Service, section 8.1.2 (page shows "Last updated: January 23rd, 2026") | "You acknowledge that the Services may generate Output containing incorrect, biased, or incomplete information." |
| Same section | "You should not rely on the Services or any Output for advice of any kind, including medical, legal, investment, financial or other professional advice." |
| Help center, reporting article | "Use the flag icon below the answer to report issues." It lists "Misinformation" among its examples. |
A help article and a legal disclaimer are different kinds of text, written for different purposes, so this page lists each as printed and does not treat them as a contradiction. Both Terms sentences sit in section 8.1.2.
How does Perplexity's documentation describe its crawlers?
It names two user agents, PerplexityBot and Perplexity-User, and treats robots.txt differently for each. This section reports only what the documentation states. The robots.txt file is, in this site's plain explanation, the public text file on your site that tells crawlers which pages they may read.
| User agent | What Perplexity's crawler documentation says it is for | robots.txt, as Perplexity states it |
|---|---|---|
| PerplexityBot | "PerplexityBot is designed to surface and link websites in search results on Perplexity. It is not used to crawl content for AI foundation models." | Help article: "Perplexity respects robots.txt directives." and PerplexityBot "will not index the full or partial text content of any site that disallows it via robots.txt." It adds: "However, if a page is blocked, we may still index the domain, headline, and a brief factual summary." |
| Perplexity-User | "Perplexity-User supports user actions within Perplexity. When users ask Perplexity a question, it might visit a web page to help provide an accurate answer and include a link to the page in its response." | "Since a user requested the fetch, this fetcher generally ignores robots.txt rules." |
The crawler page recommends: "To ensure your site appears in search results, we recommend allowing PerplexityBot in your site’s robots.txt file and permitting requests from our published IP ranges listed below." The help article says "Perplexity does not build foundation models, so your content will not be used for AI model pre-training."
The robots.txt help article (last modified September 14, 2026) says: "Previously, users could prompt Perplexity to summarize a specific URL, even if it was blocked by robots.txt." It says "this feature has been disabled to prevent misuse" and then "Now, PerplexityBot only crawls content in compliance with robots.txt." That article does not name Perplexity-User and the crawler page does not mention the change, so the pages do not show how the statements relate. This page reports both as written and says nothing about what the crawlers do beyond these documents.
The "no training" statements here concern pages your website publishes. The data section below concerns what you type into your own account. They are separate.
What can you check on your own site?
Check your robots.txt and your firewall for the two names. These steps are this site's suggestions; no page we read promises a citation.
- Open yourdomain.com/robots.txt in a browser. Look for lines that name PerplexityBot or Perplexity-User, and for a blanket block aimed at all bots. A blanket block would include PerplexityBot; that is this site's explanation of how robots.txt works, not a Perplexity statement.
- If a security service or firewall sits in front of your site, ask whoever manages it. The crawler page says: "you may need to explicitly whitelist Perplexity’s bots to ensure they can access your content."
- After any change, wait: "it may take up to 24 hours for our systems to reflect changes."
- Decide what you want. Allowing PerplexityBot is what the crawler documentation recommends "to ensure your site appears in search results" (Perplexity's wording); disallowing it is, per the help article, how a site keeps its full text out of the index.
See also the small business AI automation guide and AI agents for small businesses. This site writes pages meant to be cited by tools like this one; no result is promised.
What do the plans include, and what do the pages print as prices?
There is a free tier, and paid prices appear on several Perplexity pages, each with its own billing wording. Pages read October 7, 2026. The pages show a dollar sign without naming a currency, so no currency is stated here.
| Plan | As Perplexity's page prints it | Page |
|---|---|---|
| Free | "Standard (Free)"; "Practically unlimited basic searches"; "Very limited amount of Pro Searches" | Plans help article, last modified October 2, 2026 |
| Pro | "$17 /month or equivalent, when billed annually"; a footnote under the Pro and Max cards on the same page says "For personal use only and subject to our policies." | Pro page |
| Enterprise Pro | "Per-seat pricing (starts at $40/month or $400/year/seat)" | Plans help article, section on Enterprise Pro |
| Enterprise tab, first card (its heading is a logo image whose accessible label is "Enterprise Pro") | "$34 / month per seat when billed annually" | Enterprise pricing page |
Two cautions. First, the footnote "For personal use only" sits below the plan cards rather than inside the Pro card, and it does not say whether a one-person business counts, so ask Perplexity before buying Pro for business work. Second, the two Enterprise Pro figures are printed on different pages with different billing terms: one with a monthly and a yearly seat price and the words "starts at", the other as a monthly price when billed annually. Neither page explains how they relate; confirm the seat price with Perplexity before budgeting.
What happens to your data, and does it train AI?
On Free, Pro and Max, Perplexity's data page says "AI Data Retention is enabled by default" and that you can control whether your data is used for AI model training; on Enterprise it says "Perplexity Enterprise data is never used for AI training purposes." The page, last modified September 3, 2026, names the setting "AI data retention" and gives this path: Account settings, then Preferences, then the AI data retention toggle.
The page also states:
- "If the toggle is enabled, your data is being collected for AI training."
- "Opt-outs only apply to data collected after the opt-out date."
- "Previously collected training data cannot be deleted or removed."
- Downgrading: "If you choose to downgrade to a Free or non-Enterprise plan, your data would then follow the standard consumer data policy, where AI data retention is enabled by default."
The model-providers help page (last modified September 14, 2026) is about the providers' training, not Perplexity's own: "Perplexity’s agreements with third-party model providers like OpenAI and Anthropic prohibit using Perplexity data for training their models."
FAQ
Is Perplexity AI free?
There is a free tier. As of October 7, 2026, Perplexity's plans page lists "Standard (Free)" with "Practically unlimited basic searches" and a "Very limited amount of Pro Searches"; paid plans are on the same page.
Does Perplexity use my searches to train AI?
For Free, Pro and Max users, Perplexity's data page says AI Data Retention is enabled by default and that these users can control whether their data is used for AI model training in Account settings. It says Enterprise data is never used for AI training purposes. The setting is named "AI data retention", under Account settings, then Preferences; check the toggle in your own account.
Does Perplexity follow robots.txt?
Perplexity's help article says PerplexityBot will not index the full or partial text content of any site that disallows it via robots.txt. Its crawler documentation says of Perplexity-User, which supports user actions and may visit a web page when a user asks a question, that it generally ignores robots.txt rules because a user requested the fetch.
Is Perplexity AI accurate?
Perplexity's pages give no accuracy rate. The help center encourages double-checking sources, and the Terms of Service say the Services may generate Output containing incorrect, biased, or incomplete information.
How this page was made
Drafted with Claude (Anthropic, a competitor of Perplexity's in AI assistants). Perplexity's help center says its platform integrates AI models "including OpenAI’s GPT models and Anthropic’s Claude"; this page compares no other assistant.
It was checked against each vendor's published material on October 7, 2026. Every statement about Perplexity comes from its own help center, pricing pages, Terms of Service and crawler documentation, read in a browser from outside the United States; no third-party source was used. This page does not cover legal disputes, press reports, usage or traffic figures, or accuracy rates for Perplexity, and tested nothing in Perplexity itself. Plans, prices and settings change often; check the linked pages.