Moonshot AI Unveils Trillion-Parameter Model Kimi K3

Moonshot AI released the Kimi K3 model, featuring 2.8 trillion parameters and 1 million context length. The technical report emphasizes sparse routing and Delta Attention over simply scaling dense models.

2026-07-25 ~ 2026-07-26 · 2 related posts

• Moonshot AI launches Kimi K3 with 2.8T parameters and a 1M-token context — dl_weekly · 2026-07-25
• Moonshot’s Kimi K3 report says sparse MoE and delta attention beat bigger dense models — SumitGup · 2026-07-26