Skip to content
The blog

Written after the deployment, not before.

Essays on AI architecture, strategy and the org problems in between. Heart what’s useful, argue with me in the comments.

Tagged open-weightsclear ✕


Jul 28, 2026

open-weights5 min read

Kimi K3: a 2.8-trillion-parameter open model that bets on linear attention

Moonshot's new open-weight flagship isn't chasing a leaderboard — it's a bet that a cheap million-token context changes what we build. What actually changed under the hood, how to run it, and where it sits against the open-weight frontier.

Modernist title card reading Kimi K3 — 2.8T open weights, 1M context, linear attention.

3 reads1 hearts1 comments