Moonshot AI is open-sourcing Kimi K2.6, a model with state-of-the-art coding, long-horizon execution, and agent swarm capabilities. It is available via Kimi.com, Kimi App, API, and Kimi Code. The model shows significant improvements over K2.5 in long-horizon coding tasks, as demonstrated by internal benchmarks and enterprise beta tests.

Open-weight release

How it measures

Artificial Analysis · Intelligence Index · Reasoning

27147thof 696 models

Coding index61.8 · 74th

Median 12.5Top 57.6

Benchmarks, as Artificial Analysis ran them

91%GPQA Diamond
38%Humanity's Last Exam
1%Terminal-Bench 4.0
66%Terminal-Bench 2.1
44%Terminal-Bench Hard
52%SciCode
96%τ²-Bench
23%τ²-Bench Banking
76%IFBench
81%Long context reasoning

Per 1M tokens: $0.95 in · $4.00 out
Context: 262K tokens

LMArena · Text59thof 416· 10th of 226 open models

1Gemini 4 Argon (High)
2Muse Wonder (Max)
3Claude Opus 5.5 (High)
4Claude Fable 5 (High)

kimi-k2.6 · 1,461 (1,456–1,465) · 40,013 votes

Rated by versionKimiKimi K3 rates 26 points higher.

Kimi K2.6 · 1,461 (1,456–1,465) · 59th

Against list priceRating for the price

Kimi K2.6 · 1,461 · $1.34 per 1M tokens ($0.65 in, $3.41 out) · 262K context

LMArena, CC BY 4.0, 15 October 2026 · prices OpenRouter, list price blended 3:1, read 20 October 2026 · Artificial Analysis, read 20 October 2026

About the model

Lab: Moonshot AI
Weights: Open
Released: 21 April 2026, per models.dev
OpenRouter: Listed 20 April 2026moonshotai/kimi-k2.6

Ask your AI

Your brief, straight from the record

Connect once, then ask Claude or ChatGPT what changed for the tools you use. Every item links to its page on forck.live.

+ any MCP app

Connect your AI →