# PerplexityBot

## Quick answer

PerplexityBot is Perplexity's web crawler, used to surface and link websites inside Perplexity's search results and AI-generated answers. Unlike some AI crawlers, it is not used to train AI foundation models; it only supports Perplexity's live search and answer product, and it respects robots.txt.

## What it does

PerplexityBot crawls the web specifically to find and index content that Perplexity can surface and link to within its search answers. Perplexity states this bot is not used to crawl content for AI foundation model training, which separates its purpose from crawlers like GPTBot or ClaudeBot that explicitly serve training pipelines. PerplexityBot respects robots.txt directives, so a site can control whether it's crawled at all. It's also distinct from Perplexity-User, a separate bot that handles fetches triggered directly by a user's action inside Perplexity, such as asking it to open a specific link, rather than general indexing.

## Facts

- Operator: Perplexity
- Robots token: PerplexityBot
- User agent string(s): Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot)
- Purpose: search-index
- Respects robots.txt: yes
- IP ranges: https://www.perplexity.com/perplexitybot.json

## Should you block it?

Because PerplexityBot's stated purpose is indexing for search answers rather than model training, blocking it has a more direct effect on AI visibility than blocking a training-only bot does: a site that disallows PerplexityBot risks not being surfaced or linked when Perplexity answers a related question. Perplexity itself recommends allowing the bot specifically so a site can appear in its results, the opposite tradeoff from training-focused crawlers where blocking mainly protects content from reuse. A brand trying to get mentioned when people ask Perplexity about its category should generally allow PerplexityBot, reserving any blocking for specific sections like paywalled or private content rather than the whole site. The content-use concern here is smaller than with training crawlers, since indexed content is being cited and linked rather than absorbed into a model's weights.

## How to block it

To allow PerplexityBot, which Perplexity recommends for appearing in its answers, no robots.txt entry is needed, or it can be stated explicitly:

User-agent: PerplexityBot
Allow: /

To block it instead:

User-agent: PerplexityBot
Disallow: /

## FAQ

### Does PerplexityBot train Perplexity's AI models?

No. Perplexity states this bot surfaces and links websites in search results and is not used to crawl content for AI foundation models.

### Should I block PerplexityBot if I want to appear in Perplexity's answers?

No. Perplexity recommends allowing it specifically so a site can be surfaced and linked in its search results.

### Is PerplexityBot the same as Perplexity-User?

No. Perplexity-User handles fetches triggered by a specific user action inside Perplexity, while PerplexityBot handles general crawling and indexing.

## Sources

- PerplexityBot surfaces and links websites in Perplexity's search results and is not used to crawl content for AI model training. Source: https://docs.perplexity.ai/guides/bots