MistralAI-Training
Collects web content to build datasets for training Mistral's generative AI models. Mistral says it isn't used for search indexing or live answers.
What happens if you block it
Mistral won't use your content for training. Mistral search and on-request fetches are controlled separately by MistralAI-Index and MistralAI-User.
Our take
Block it if you don't want your content in Mistral's training data; it won't affect your visibility.
robots.txt
To block it:
User-agent: MistralAI-Training Disallow: /
To allow it explicitly (useful when a wider rule blocks other bots):
User-agent: MistralAI-Training Allow: /
The operator's own words
MistralAI-Training crawls web content to help build datasets for training Mistral generative AI models. Webmasters can disallow this user agent in their robots.txt file.Mistral AI documentation, checked 2026-11-04
Other Mistral AI agents
MistralAI-User
Mistral AI
Fetches a page when someone asks Mistral's assistant about it, and links the page as a source.
MistralAI-Index
Mistral AI
Indexes pages for Mistral search, which answers questions in its Vibe assistant.
Can MistralAI-Training reach your site?
Free, no sign-up. We request your homepage as 18 different clients and show what each one gets.