Yandex's AI progress and the Mistral comparison
A user describes Yandex's model as a modern, small mixture-of-experts system with a Qwen 3 Next-like design.
TLDR
A user praises Yandex's small mixture-of-experts model, highlighting a Qwen 3 Next-like design and citing roughly 18 trillion tokens. They estimate Yandex is 1.5 years behind as an AGI lab and put it roughly on par with Mistral.
Combined views
31.1K
3 Sources, first seen ago
514 likes26 comments103 saves14 reposts
