Yandex Drops AliceAI Foundation 80B Open Source Model That Beats Rivals on Math

Yandex open-sourced an 80B mixture-of-experts model with 3B active parameters, 256K context, and a hybrid KDA-attention architecture trained from scratch.

$Yandex Drops AliceAI Foundation 80B Open Source Model That Beats Rivals on Math$

Source

AlphaSignal

Published

Sep 23, 2026, 12:55 AM

Author

AlphaSignal Newsroom

Read

1 min read

Yandex open-sourced an 80B mixture-of-experts model with 3B active parameters, 256K context, and a hybrid KDA-attention architecture trained from scratch.

Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.

Read original report