Moonshot AI releases Kimi K2.6
AIMade by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made
Released Kimi K2.6 — 1T total / 32B active, up to 13 hours of continuous coding, 300 subagents collaborating across 4,000 steps. SWE-Bench Verified 80.2% (self-reported)
Why it matters
An open-weight model entered the 80% range on SWE-Bench Verified. No independent party checking the result is named
Source — Primary sources first; secondary reports are labelled “Press”
Pressyicaiglobal.comhttps://www.yicaiglobal.com/news/chinas-moonshot-ai-releases-kimi-k26-pushing-boundaries-in-coding-multi-agent-capabilitiesAdded in Week 35, 2026 · See that week
On the map — ⑨ Open-Weight Lineage line
- Previous stopKimi K2 Thinking
- This recordKimi K2.6
- Next stopK-EXAONE 2.0
In the model family tree — Kimi family · 8 versions
- Earlier versionKimi K2.5
- This recordKimi K2.6
- Later versionKimi K2.7-Code