Claude-User
Fires when a Claude user asks about a page and Claude fetches it live during the conversation. Like ChatGPT-User, every hit is a real person being read your content by an assistant - a demand signal, not a crawl.
What it feeds
How to identify it
The token appears as Claude-User with a version and an Anthropic contact address. Anthropic publishes the address ranges its bots fetch from at claude.com/crawling/bots.json; the user agent alone proves nothing, because any client can send any name.
Blocking and its consequences
Anthropic states that its bots honour industry standard robots.txt directives,[3] so a disallow here does take effect - and takes the site out of live answers for users who asked about it directly. That is a different trade from blocking ClaudeBot, which only touches training.
Verification
Operator documentation: https://support.claude.com/en/articles/8896518
A user agent is a claim, not a proof: any client can send any name. Where an operator publishes IP ranges, a reverse lookup is the only way to tell a real visit from an impersonation.
Seen by Baseline
Requests from this bot across sites tracked by Baseline, last 30 days. Aggregated over - sites, never reported per site.
References (3)
Last updated 2026-09-05. Written and maintained by Baseline Labs.
