Perplexity-User: definition and how it treats robots.txt

Perplexity-User is the user agent string Perplexity sends when it fetches a page because a user's question or action called for it. It is an answer-time fetcher, not an index-time crawler.

Perplexity runs two named agents:

• PerplexityBot crawls the web for Perplexity's search index. Perplexity states it is "not used to crawl content for AI foundation models."
• Perplexity-User fetches pages on demand when a user's request triggers it. Perplexity states that "this fetcher generally ignores robots.txt rules."

For the full breakdown of both agents, see PerplexityBot vs Perplexity-User. This entry stays a tight definition.

Crawlers split into two families. Index-time crawlers fetch pages ahead of time and store them. Answer-time fetchers fetch a page in the moment, because a person asked for something. The difference is explained in index-time crawlers vs answer-time fetchers.

Perplexity-User sits in the second family. Robots.txt, as defined in RFC 9309, is a request that well-behaved crawlers choose to honor. It is not access control. Perplexity treats a user-triggered fetch as acting for that user, so it does not apply robots.txt to it in the usual way.

Perplexity is not alone in this. OpenAI says robots.txt rules "may not apply" to ChatGPT-User, its own user-initiated agent. As of October 2026, answer-time fetchers across vendors are the category where robots.txt is weakest. Anthropic's equivalent is covered in ClaudeBot, Claude-SearchBot and Claude-User explained.

The practical consequence runs in two directions.

If you want Perplexity to see you, Perplexity-User is not the agent to worry about in robots.txt. PerplexityBot is. Allowing PerplexityBot makes your pages eligible for Perplexity's index. Eligibility is not a guarantee of retrieval or citation. Nobody outside Perplexity controls which pages it selects for a given answer.

If you want to keep Perplexity out, a robots.txt rule will not reliably stop Perplexity-User. Any control has to live at a layer the fetcher cannot opt out of, such as your server, CDN or firewall. Cloudflare's bot documentation describes how verified bots are identified and managed at that layer.

The network layer cuts both ways. A firewall can also block an AI agent your robots.txt allows, and nothing in your analytics will say so. The first thing to check is what each Perplexity agent actually receives when it requests your pages. The AI Crawler Checker requests a page as each AI crawler and shows which ones your firewall turns away.

• "Disallowing Perplexity-User in robots.txt blocks it." Perplexity's documentation says the fetcher generally ignores robots.txt. The rule may be read by nobody.
• "Blocking PerplexityBot also blocks Perplexity-User." They are separate agents with separate behavior. One rule does not cover the other.
• "Allowing Perplexity-User gets you cited." No first-party source says that. Index-time access through PerplexityBot is the documented route into Perplexity's index, and even that buys eligibility only.
• "Perplexity-User visits mean Perplexity is indexing you." A user-triggered fetch reflects a request in the moment. It says nothing documented about index inclusion.

For the broader question of compliance, see does Perplexity respect robots.txt.

• PerplexityBot: Perplexity's index-time crawler. See Perplexity's crawler documentation, digested.
• Perplexity's index: whether Perplexity keeps its own index is covered in does Perplexity have its own search index.
• ChatGPT-User and Claude-User: the OpenAI and Anthropic answer-time equivalents.
• User-triggered AI fetchers: the wider category, covered in SEO for AI agents.

Is Perplexity-User the same as PerplexityBot? No. PerplexityBot crawls the web ahead of time to build Perplexity's search index, and it follows robots.txt rules. Perplexity-User fetches a page in the moment because a user's request triggered it, and Perplexity documents that it generally ignores robots.txt. Treat them as two agents with two separate controls, and check each one on its own.

How can I stop Perplexity-User from fetching my pages? Robots.txt is not a reliable tool for this, because Perplexity says the fetcher generally ignores it. Any dependable control sits at your server, CDN or firewall, where requests are accepted or refused regardless of robots.txt. Before you block anything, weigh the cost: refusing user-triggered fetches may also stop Perplexity from reading your page when someone asks about it.

Should I add a Perplexity-User line to robots.txt at all? It does no harm, and it records your intent. Just do not rely on it as enforcement, since Perplexity documents that the fetcher generally ignores robots.txt. The line that matters more for Perplexity visibility is the one for PerplexityBot. If it is disallowed, directly or through a blanket AI block, your pages are not eligible for Perplexity's index.

Does allowing Perplexity-User help me get cited in Perplexity? There is no first-party evidence that it does. The documented route into Perplexity's answers runs through PerplexityBot and Perplexity's index, and allowing that crawler buys eligibility only. Which pages Perplexity retrieves and cites for a given question is decided by Perplexity, not by the publisher. Nobody can promise a citation from either agent.