Anthropic’s three crawlers explained
Blocking ClaudeBot doesn’t remove you from Claude’s search answers anymore, and blocking Claude-SearchBot can. Here is how Anthropic’s bots split up.
Anthropic’s crawler documentation lists three separate bots, each with its own user agent you can allow or block in robots.txt. Reports date the documentation update to late February 2026. The previous version described only ClaudeBot.
What happened
- ClaudeBot collects public web content for training Anthropic’s models.
- Claude-SearchBot crawls to index content for search, so Claude can find and cite pages in answers. Anthropic warns that blocking it may reduce your visibility and accuracy in Claude’s search results.
- Claude-User fetches a page when a user asks Claude about it.
Older user agents, Claude-Web and anthropic-ai, are deprecated. Anthropic publishes the IP ranges its bots use, so you can verify that a request claiming to be Claude really comes from Anthropic.
What it means for you
- Block training, keep search: disallow
ClaudeBotand allowClaude-SearchBotandClaude-User. - Check old rules. A robots.txt that only blocks
anthropic-aidoes nothing today. - Verify before you trust. Scrapers often pretend to be ClaudeBot. Compare request IPs with Anthropic’s published ranges.
The AI Bot IP Checker tells you whether an IP really belongs to the bot it claims to be.