PerplexityBot
Builds the index Perplexity's answer cards draw from. Pages crawled here are eligible to appear as cited sources. Perplexity has publicly stated this bot is not used to collect training data for foundation models.
Specs
| Vendor | Perplexity AI |
| Category | SEARCH |
| robots.txt token | PerplexityBot |
| Renders JavaScript | HTTP only |
| Honors robots.txt | yes |
| Reverse-DNS pattern | *.perplexity.ai |
| IP-range source | https://www.perplexity.ai/perplexitybot.json |
User-Agent string
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; PerplexityBot/1.0; +https://perplexity.ai/perplexitybot)Considerations
- 2024 controversy: Perplexity was reported by Wired to fetch content from sites that had disallowed PerplexityBot in robots.txt, by routing through residential IPs with a generic Chrome UA. The company has since updated its policy and the docs reflect that.
- HTTP-only for indexing. Live retrieval is handled by the separate Perplexity-User token.
robots.txt recipe
User-agent: PerplexityBot
Allow: /Sources: Perplexity · Bots · Wired · Perplexity robots.txt report (Aug 2024)