Cohere-Command
Cohere-Command is a live-fetch agent operated by Cohere. It does not crawl the web on a schedule. It hits your site only when an end-user asks the underlying AI a question that requires fresh information from a specific page.
Traffic is bursty and unpredictable. A single trending topic can send hundreds of Cohere-Command requests in an hour, then nothing for days. Each request typically reads one or two pages, not your whole site.
Allowing Cohere-Command is how your content becomes part of Cohere's answers. Blocking it means users asking that AI about your topic will be answered using someone else's content instead.
See Cohere-Command on your own site
Match the User-Agent header on incoming requests against the pattern below.
regex
For higher confidence, also verify the source IP against the operator's published ranges. UA strings can be spoofed; IP ownership is harder to fake.
Renders JavaScript
Sometimes
IP verification
Published IP ranges
Crawl frequency
Burst, user-driven
Honors robots.txt
Yes
Honors Crawl-delay
Varies
Cohere runs 4 bots in total. Each one is a separate user-agent so you can allow or block them independently.
Training Crawler
2Agentic Browser
1Live-Fetch AI
1- Cohere-CommandYou are here
Should I let Cohere-Command through?
In most cases, yes. Live-fetch agents drive citations inside AI answers. Allowing keeps your content in the conversation. If volume gets noisy, rate-limit it before you block it outright.
Does blocking Cohere-Command affect my Google rankings?
No. Cohere-Command fetches a page only when a user is actively asking Cohere a question. It has nothing to do with how Google or Bing rank you. The cost of blocking is that Cohere can't quote your content in its answer.
How do I confirm a request is really from Cohere-Command?
Look at the User-Agent header in your access logs and match it against the strings listed above. Worth knowing that the User-Agent is easy to fake, so this check tells you "the traffic claims to be Cohere-Command", not "the traffic is genuinely Cohere-Command". If you need stronger guarantees, look for a reverse-DNS check or wait for Cohere to publish IP ranges.
Does a Cohere-Command visit count as a real user visit?
Sort of. There is a human asking Cohere a question on the other end, but they never load your page in their own browser. They see whatever Cohere quotes back, usually a snippet plus a citation link. Count it as upstream attention rather than as a session.
How is Cohere-Command different from Cohere's other bots?
Cohere splits work across multiple user-agents so site owners can decide on each one independently. Training crawlers, live-fetch agents, search indexers, and agentic browsers each get their own name. Worth scanning the rest of the Cohere family above to see which ones actually matter for your site.
What's the cleanest way to control Cohere-Command?
Two layers. Robots.txt for the polite crawlers that read it, and rules at your CDN or edge for the ones that don't. Rankly's Agent Experience handles both from a single config, so you can allow, block, rate-limit, or serve a stripped-down version per bot. Agent Analytics handles the observation half so you know which bots are actually worth a rule.
Verify everything above against the operator's own documentation.