Skip to content
Primetra
All agents

Agent reference · Training crawler

anthropic-ai (legacy)

A legacy Anthropic name that may still appear. Anthropic's current crawlers are ClaudeBot, Claude-User and Claude-SearchBot.

Vendor
Anthropic
Agent type
Training crawlerCollect pages to train AI models. Blocking one keeps your pages out of future training, not out of answers.
Can be verified by
Published addressesThe vendor publishes where this agent runs from, so a visit can be checked against that list.
How sure we are
DocumentedThe vendor's own documentation describes this agent.
Last reviewed
Vendor documentation
support.claude.com (opens in a new tab)

Tokens to look for

User-Agent tokens

Appear in the User-Agent header of its requests.

  • anthropic-ai

robots.txt tokens

No robots.txt token is documented for it.

Check a visitor claiming to be anthropic-ai (legacy)

We've filled in its token. Paste the full User-Agent from your logs and the address it came from for a real answer.

Paste the full User-Agent header from your logs. A single token such as GPTBot works too.

Optional

IPv4 or IPv6, as your logs show it. Used only for this check and never stored. Without it, a claim can only come back as can't be verified.

What this means for your site

anthropic-ai (legacy) gathers pages for model training. Allowing it is a choice about future models, not about whether assistants can cite you today: that is the search crawlers' job.

See whether AI agents can find your product, use it and, soon, pay for it.

Run the free agent-readiness check