Skip to content

cohere-ai

User fetcher

Cohere's fetcher for pages a person asks it to read. Seen on 48 of 135 sites in 13 weeks.

cohere-ai reached 32 of 92 sites last week

Sites that got at least one cohere-ai request each week, against every site reporting that week.

Reach38%12th of 2639 of 104 sitesLast 4 weeks
Requests44K1% of all crawler requests13 weeks, 48 sites
Per site20a weekMedian where it visitsLast 4 weeks

cohere-ai reads a page when a person asks for it.

Cohere says cohere-ai follows robots.txt. A rule changes future crawls, not pages it already collected.

Block cohere-ai
User-agent: cohere-ai
Disallow: /
Allow cohere-ai
User-agent: cohere-ai
Allow: /
User agentcohere-ai

Is it really cohere-ai?

Anyone can send the cohere-ai name. Here is how to check a request that carries it.

How to check
1There is no address list to matchCohere does not publish addresses for cohere-ai, so a matching name is a clue, not proof.
2Never count a secret-file requestA request for /.env or /.git is not a crawl, whatever name it carries.
See every fake crawler

Common questions

What is cohere-ai?

cohere-ai is Cohere's user fetcher. Cohere token reported for retrieving data in response to user-initiated prompts.

How do I block cohere-ai?

Add a directive to your robots.txt: "User-agent: cohere-ai" followed by "Disallow: /". Cohere says cohere-ai follows robots.txt. A rule changes future crawls, not pages it already collected.

Does blocking cohere-ai hide me from AI?

No single bot rule hides a site from every AI product. Treat hits as retrieval for Cohere-powered products unless logs show bulk crawling behavior.