Model

Kimi K2 Thinking

1T/32B-active MoE with 256K context, native INT4 QAT, Modified MIT license; deep-thinking variant with 200-300 sequential tool calls for agentic tasks.

Made by: Moonshot AI
Released: November 6, 2025

The community’s take.

Specific experiences are worth more than sweeping claims.

1 TAKES

@hn_ac70a4

5 months ago

Public discussion · Hacker News

Public discussion, not a rating submitted on Rate My AI.

Weird, I have gone local for the last 2 years. I use Chinese models 90% of the time, Kimi K2 Thinking, DeepSeekv3.Terminus, Qwen3 and GLM4.6. I'm not vibe testing it but really putting them to use and they do keep up great.

be honest, be specific

Back to the directory Add to comparison