Skip to content

Kimi K2 Thinking

Moonshot AI · Nov 6, 2025

Thinking Kimi model for slower research passes, planning, and hard technical questions

ECI STANDING#73 of 266

SCORE145.76143.25–147.14 CI

FRONTIER LAGLaunched 6.7 months behindScores about what o3 did in April 2025.

SCORES LIKE THESE MODELSGLM-5Z.ai (Zhipu AI) · Feb 2026145.85 GPT-5.4 NanoOpenAI · Mar 2026145.85

MiniMax-M2.7MiniMax · Mar 2026145.8

o4-miniOpenAI · Apr 2025145.65

Full Epoch Capabilities Index ranking ↗

From the lab

Related articleKimi K2 Thinking

RELEASE DATENov 6, 2025

WEIGHTSOpen weights

CONTEXT WINDOW262.1K

MAX OUTPUT262.1K

REASONINGYes

TOOL CALLINGYes

INPUTStext

OUTPUTStext

Open in the timeline Specifications ↗

Specifications from Models.dev. Availability and limits may differ by provider. Benchmark scores come from Epoch AI under CC BY 4.0.