Compare AI models

Specifications
Parameters
8B/70B—
Context window
8k—
Benchmarks
HumanEval
81.7%72%
MATH
50.4%47.9%
MMLU
82%81.5%
GSM8K
93%87.6%
Overview
CompanyMetaZ.ai
Release dateApr 18 2024Jan 16 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

LLaMA 3 (8B/70B) leads GLM-4 on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). GLM-4 shipped 93 days before LLaMA 3 (8B/70B), so benchmark comparisons should account for the intervening progress.

LLaMA 3 (8B/70B) is open weight, while GLM-4 is closed.

On HumanEval, LLaMA 3 (8B/70B) leads at 81.7% vs GLM-4 at 72%. On MATH, LLaMA 3 (8B/70B) leads at 50.4% vs GLM-4 at 47.9%. On MMLU, LLaMA 3 (8B/70B) leads at 82% vs GLM-4 at 81.5%. On GSM8K, LLaMA 3 (8B/70B) leads at 93% vs GLM-4 at 87.6%.