Grok 4.6vsGLM-5.2

Grok 4.6
GLM-5.2
Specifications
Parameters
744B
Context window
1M
Benchmarks
Nonsense detection
BullshitBench v2
31%
Agentic coding
SWE-Bench Pro
62.1%
Agentic coding
CursorBench v3.2
69.9%Best
55%
Agentic coding
CursorBench v3.1
54.6%
Agentic coding
DeepSWE 1.1
65.9%Best
44%
Agentic coding
FrontierCode v1.1 (Extended) · extended split
61.3%
Expert software engineering
APEX-SWE
56.4%
Next.js coding
Next.js Evals
88%
Agentic computer work
Frontier-Bench v0.1
5.1%
Agentic terminal coding
Terminal-Bench 3.0
26%
Agentic terminal coding
Terminal-Bench 2.1
81%
Expert agentic work
APEX-Agents
57.5%
Multidisciplinary reasoning
Humanity's Last Exam · no tools
40.5%
Multidisciplinary reasoning
Humanity's Last Exam · with tools
54.7%
Science
GPQA Diamond
91.2%
Agentic legal work
Harvey's Legal Agent Benchmark
15.8%
Overall intelligence
AA Intelligence Index
61
Knowledge work
GDPval-AA v2
1753Best
1514
Knowledge work
AA-Briefcase
1577
Community preference (code)
Arena Elo (Code)
1587
Overview
CompanySpaceXAIZ.ai
Release dateAug 12 2026Jun 16 2026
AccessProprietaryOpen Weight

Which is better: Grok 4.6 or GLM-5.2?

Grok 4.6 leads GLM-5.2 on 3 of the 3 benchmarks they both report (CursorBench v3.2, DeepSWE 1.1, GDPval-AA v2). GLM-5.2 shipped 57 days before Grok 4.6, so benchmark comparisons should account for the intervening progress.

Grok 4.6 is proprietary, while GLM-5.2 is open weight.

On CursorBench v3.2, Grok 4.6 leads at 69.9% vs GLM-5.2 at 55%. On DeepSWE 1.1, Grok 4.6 leads at 65.9% vs GLM-5.2 at 44%. On GDPval-AA v2, Grok 4.6 leads at 1753 vs GLM-5.2 at 1514.

Frequently asked questions

Grok 4.6 was released by SpaceXAI on Aug 12 2026.

GLM-5.2 was released by Z.ai on Jun 16 2026.

Grok 4.6 leads on CursorBench v3.2 — Grok 4.6 69.9% vs GLM-5.2 55%.

Grok 4.6 is a proprietary model released by SpaceXAI. GLM-5.2 is an open weight model released by Z.ai.

Other comparisons