Kimi K2.7 CodevsGLM-5.3

Kimi K2.7 Code
GLM-5.3
Specifications
Parameters
1T
743B
Context window
256k
Benchmarks
Agentic coding
CursorBench v3.2
49.7%
Agentic coding
DeepSWE 1.1
31%
66.9%Best
Next.js coding
Next.js Evals
75%
Agentic terminal coding
Terminal-Bench 3.0
28.3%
Cybersecurity
CyberGym
84.5%
Cybersecurity
ExploitBench
54.4%
Cybersecurity
ExploitGym · 6-hour budget
130
Cybersecurity
ExploitGym · 2-hour budget
105
Multidisciplinary reasoning
Humanity's Last Exam · with tools
62.5%
Agentic computer use
Agent's Last Exam
28.5%
Business workflows
AutomationBench
48.2%
Knowledge work
GDPval-AA v2
1769
Community preference (code)
Arena Elo (Code)
1469
Overview
CompanyMoonshot AIZ.ai
Release dateJun 12 2026Aug 14 2026
AccessOpen WeightProprietary

Which is better: Kimi K2.7 Code or GLM-5.3?

GLM-5.3 leads Kimi K2.7 Code on 1 of the 1 benchmark they both report (DeepSWE 1.1). Kimi K2.7 Code shipped 63 days before GLM-5.3, so benchmark comparisons should account for the intervening progress.

Kimi K2.7 Code has 1T parameters, while GLM-5.3 has 743B. Kimi K2.7 Code is open weight, while GLM-5.3 is proprietary.

On DeepSWE 1.1, GLM-5.3 leads at 66.9% vs Kimi K2.7 Code at 31%.

Frequently asked questions

Kimi K2.7 Code was released by Moonshot AI on Jun 12 2026.

GLM-5.3 was released by Z.ai on Aug 14 2026.

GLM-5.3 leads on DeepSWE 1.1 — Kimi K2.7 Code 31% vs GLM-5.3 66.9%.

Kimi K2.7 Code is an open weight model released by Moonshot AI. GLM-5.3 is a proprietary model released by Z.ai.

Other comparisons