MistralAI-Training: what it is and how to block it

Operator

Mistral AI

Type

AI training

robots.txt token

MistralAI-Training

Follows robots.txt

Yes, per Mistral AI

IP ranges

Not published

Blocked by name

0.5% of top sites

What is MistralAI-Training?

MistralAI-Training is Mistral AI's AI training crawler: it collects content that may be used to train AI models. In Mistral AI's words:

“MistralAI-Training crawls web content to help build datasets for training Mistral generative AI models.”

“This crawler is not used for search indexing or to answer live user queries in Vibe.”

MistralAI-Training user agent string

Mistral AI documents this user agent string for MistralAI-Training:

Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; MistralAI-Training/1.0; +https://docs.mistral.ai/robots)

In robots.txt, use the token MistralAI-Training, not the full string. Crawlers match on the token, and version numbers in the full string can change.

How many websites block MistralAI-Training?

In our Oct 8, 2026 check of 817 of the most-visited websites, 0.5% block MistralAI-Training by name for their whole site (4 sites).

For comparison, 0.1% of the same sites block Googlebot by name. The sites are the top 1,000 origins in the Chrome UX Report for August 2026; see how we measured.

How to block MistralAI-Training in robots.txt

Add this group to the robots.txt file at the root of your domain to block MistralAI-Training from your whole site:

User-agent: MistralAI-Training
Disallow: /

To block only part of your site, list those paths instead:

User-agent: MistralAI-Training
Disallow: /private/

A group that names MistralAI-Training replaces your User-agent: * rules for it entirely (RFC 9309). If your catch-all group blocks paths that MistralAI-Training should also stay out of, repeat them in its group. To let MistralAI-Training in while a catch-all rule blocks others:

User-agent: *
Disallow: /

User-agent: MistralAI-Training
Allow: /

“Webmasters can disallow this user agent in their robots.txt file.”

What happens if you block MistralAI-Training?

Mistral AI doesn't document what blocking MistralAI-Training changes.

How to verify requests from MistralAI-Training

Mistral AI doesn't publish IP ranges or a verification method for MistralAI-Training. Any client can send its user agent string, so treat requests that claim to be MistralAI-Training as unverified.

Other Mistral AI crawlers

• MistralAI-Index · Search
• MistralAI-User · User-triggered

See every AI crawler and how often top sites block it