AI Model Release Tracker - Analytics

Claude Opus 5vsGLM-5

Claude Opus 5
GLM-5
Specifications
Parameters
744B
Context window
1M
Benchmarks
Nonsense detection
BullshitBench v2
28%
Coding
SWE-Bench Verified
77.8%
Multilingual coding
SWE-Bench Multilingual
73.3%
Agentic coding
DeepSWE 1.1
68.8%
Agentic coding
FrontierCode v1.1 (Main)
53.4%
Agentic computer work
Frontier-Bench v0.1
43.3%
Agentic terminal coding
Terminal-Bench 2.0
56.2%
Web browsing
BrowseComp
90.8%Best
75.9%
Multidisciplinary reasoning
Humanity's Last Exam · no tools
56.3%
Multidisciplinary reasoning
Humanity's Last Exam · with tools
64.7%Best
50.4%
Novel problem-solving
ARC-AGI-3
30.2%
Biology
BioMysteryBench · hard
49.4%
Biology
BioMysteryBench · human solved
90.1%
Science
GPQA Diamond
86%
Agentic computer use
OSWorld 2.0
70.6%
Business workflows
AutomationBench
26%
Agentic legal work
Harvey's Legal Agent Benchmark (Held-out)
11.7%
Health
HealthBench Professional
59.8%
Knowledge work
GDPval-AA v2
1861
Community preference (code)
Arena Elo (Code)
1430
Overview
CompanyAnthropicZ.ai
Release dateJul 24 2026Feb 12 2026
AccessProprietaryOpen Weight

Which is better: Claude Opus 5 or GLM-5?

Claude Opus 5 leads GLM-5 on 2 of the 2 benchmarks they both report (BrowseComp, Humanity's Last Exam). GLM-5 shipped 162 days before Claude Opus 5, so benchmark comparisons should account for the intervening progress.

Claude Opus 5 is proprietary, while GLM-5 is open weight.

On BrowseComp, Claude Opus 5 leads at 90.8% vs GLM-5 at 75.9%. On Humanity's Last Exam · with tools, Claude Opus 5 leads at 64.7% vs GLM-5 at 50.4%.

Frequently asked questions

Claude Opus 5 was released by Anthropic on Jul 24 2026.

GLM-5 was released by Z.ai on Feb 12 2026.

Claude Opus 5 leads on Humanity's Last Exam · with tools — Claude Opus 5 64.7% vs GLM-5 50.4%.

Claude Opus 5 is a proprietary model released by Anthropic. GLM-5 is an open weight model released by Z.ai.

Other comparisons