LLaMA 3.1vsGLM-5.3

LLaMA 3.1
GLM-5.3
Specifications
Parameters
743B
API pricing
Cheapest input
$0.02DeepInfra$1.40Z.AI
Cheapest output
$0.04DeepInfra$4.40Z.AI
Benchmarks
BullshitBench v2
14%71%
Benchmarks
DeepSWE 1.1
66.9%
Terminal-Bench 3.0
28.3%
CyberGym
84.5%
ExploitBench
54.4%
ExploitGym · 6-hour budget
130
ExploitGym · 2-hour budget
105
Humanity's Last Exam · with tools
62.5%
Agent's Last Exam · pass@1
28.5%
AutomationBench
48.2%
GDPval-AA v2
1769
Overview
CompanyMetaZ.ai
Release dateJul 23 2024Aug 14 2026
AccessOpen WeightProprietary

Other comparisons

LLaMA 3.1vsClaude Opus 5GLM-5.3vsClaude Opus 5LLaMA 3.1vsGPT-5.6 SolGLM-5.3vsGPT-5.6 SolLLaMA 3.1vsGemini 3.7 FlashGLM-5.3vsGemini 3.7 FlashLLaMA 3.1vsGrok 4.6GLM-5.3vsGrok 4.6LLaMA 3.1vsDeepSeek-V4-Pro-0813GLM-5.3vsDeepSeek-V4-Pro-0813LLaMA 3.1vsMistral Medium 3.5GLM-5.3vsMistral Medium 3.5

Frequently asked questions

GLM-5.3 leads LLaMA 3.1 on 1 of the 1 benchmark they both report (BullshitBench v2). LLaMA 3.1 shipped 752 days before GLM-5.3, so benchmark comparisons should account for the intervening progress.

LLaMA 3.1 is open weight, while GLM-5.3 is proprietary.

On BullshitBench v2, GLM-5.3 leads at 71% vs LLaMA 3.1 at 14%.

LLaMA 3.1 was released by Meta on Jul 23 2024.

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

LLaMA 3.1 is an open weight model released by Meta. GLM-5.3 is a proprietary model released by Z.ai.