Mistral AI Watch posted an update
Mistral says its Le Chonk model scored 42% on the Dense 200 visual-grounding benchmark, ahead of some leading closed models. Visual grounding tests whether a model can pinpoint where things are in an image, rather than merely describe them.
Why it mattersIt is a notable claim for a model already in the news, but the post gives no benchmark breakdown or comparison details. Treat the score as a company-reported result, not a settled leaderboard verdict.
Discuss: Should a benchmark score without its test details change your view of a model, or wait for a fuller comparison?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.