Perplexity uses two bots:
- PerplexityBot crawls the web so Perplexity can find and link to websites in its answers. Perplexity says it is not used to train AI models.
- Perplexity-User fetches a page when a person asks Perplexity something that needs it.
Each has its own official IP list, so both can be verified.
Side by side
| Bot | What it does | robots.txt | Official IP list |
|---|---|---|---|
| PerplexityBot | Indexes pages so Perplexity can surface and cite them | Respects it | perplexity.ai/perplexitybot.json |
| Perplexity-User | Fetches a page for a person’s question, right now | Generally doesn’t apply, Perplexity says, because a person asked | perplexity.ai/perplexity-user.json |
PerplexityBot: being found in Perplexity
If you want your pages cited as sources in Perplexity’s answers, allow PerplexityBot. Blocking it tells Perplexity not to index your site, which makes it much less likely that your pages show up as sources.
Its user-agent contains PerplexityBot:
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot)Perplexity-User: a person is reading your page through Perplexity
When someone asks Perplexity a question and it opens your page to answer, the request is made by Perplexity-User. Perplexity’s documentation says that, because a user requested the fetch, this fetcher generally ignores robots.txt rules. If you don’t want these fetches, you need to block them at the server or firewall, using Perplexity’s published list for Perplexity-User.
These visits are worth counting rather than blocking for most sites: each one is a person who is reading about your page, right now.
How to verify them
Check the address against the list for that exact bot. A request that says “PerplexityBot” but comes from an address outside perplexitybot.json is a fake. Fakes using the PerplexityBot name are common, because scrapers know many sites let AI search bots in.
- See the current ranges on the PerplexityBot and Perplexity-User pages.
- Paste any visit into the bot checker.
- Read how to verify any AI crawler for the method.
Agents that don’t say they’re a bot
AI tools increasingly browse with an ordinary browser user-agent: Perplexity’s Comet browser, for example, looks like Chrome. Such visits don’t name a bot at all, so a user-agent rule can’t see them. When they come from a company’s published addresses, an address check still can. OZMAICO checks every visit’s address against the published AI lists for exactly this reason.
robots.txt examples
Be cited by Perplexity (the default if you don’t mention it):
User-agent: PerplexityBot
Allow: /Ask Perplexity not to index your site:
User-agent: PerplexityBot
Disallow: /Remember that this doesn’t stop Perplexity-User, and doesn’t stop fakes. See robots.txt rules for AI bots and fake AI bots.
People who come from Perplexity
Visitors who click a source in a Perplexity answer arrive with perplexity.ai as the referrer. Perplexity shows its sources next to every answer, so being one of them can bring real visitors. Count them with an “AI assistants” channel in your analytics; see how to find AI traffic in your analytics.
See Perplexity on your site
Create a free account to see verified PerplexityBot and Perplexity-User visits, the fakes, the pages they read, and the people Perplexity sends you.
Check your own website
See which AI bots really visit your site — real or fake, page by page — and how many people AI assistants send you. Free.