Moonshot AILaunchAI ModelsNovember 3, 2026

Moonshot AI releases Kimi K3.1 Pro open weights

Read original source →huggingface.co

Kimi K3.1 Pro is a mixture-of-experts model with 2.8T total parameters (~104B activated per inference), a ~1M-token context window, and weights around 1.5TB. Moonshot calls it its most capable 3T-class model; Nebius says coding performance is close to Opus 5.5 on Terminal-Bench 4.0.

How this story unfolded

11 days · 2 reports · 13 community posts · 15 of 18 shown

  1. Oct 25
  2. Nov 4
  3. Nov 5

More stories today

Open the live feed