Kimi K2 0905

Creator

Kimi

Released

2025-09-05

Intelligence

23.5

Artificial Analysis Index

Coding

Artificial Analysis Index

In $/1M

$0.60

input tokens

Out $/1M

$2.50

output tokens

Blended $/1M

$1.07

3:1 blended

Speed

0

tokens / sec

Profile

License, openness, and modality — the governance layer.

License

Proprietary

non-commercial

Weights

Closed

API only

Modalities

Text

Parameters

1T

total

Kimi K2 0905 is the September update of Kimi K2 0711. It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It supports long-context inference up to 256k tokens, extended from the previous 128k. This update improves agentic coding with higher accuracy and better generalization across scaffolds, and enhances frontend coding with more aesthetic and functional outputs for web, 3D, and related tasks. The model is trained with a novel stack incorporating the MuonClip optimizer for stable large-scale MoE training.

Capability profile

Category strength across 20 domains, via LLM Stats.

code
90legal
90finance
90language
90healthcare
90math
80
biology
80general
80physics
80chemistry
80reasoning
80
Via LLM Stats

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
77
Humanity’s Last Exam
6
MMLU-Pro
82
SciCode
31
LiveCodeBench
61
MATH-500
AIME 2025
57
τ²-Bench (agentic)
73
Terminal-Bench Hard
23
IFBench
42
Via Artificial Analysis
© 2026 NYSGPT2525 LLC