Compare AI models

Specifications
Parameters
8B/70B—
Context window
8k128k
Benchmarks
HumanEval
81.7%75.7%
MATH
50.4%57.9%
MMLU
82%81.9%
GSM8K
93%90.9%
Overview
CompanyMetaZ.ai
Release dateApr 18 2024Jun 5 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

LLaMA 3 (8B/70B) leads GLM-4-Air on 3 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). LLaMA 3 (8B/70B) shipped 48 days before GLM-4-Air, so benchmark comparisons should account for the intervening progress.

Context windows are 8k (LLaMA 3 (8B/70B)) vs 128k (GLM-4-Air). LLaMA 3 (8B/70B) is open weight, while GLM-4-Air is closed.

On HumanEval, LLaMA 3 (8B/70B) leads at 81.7% vs GLM-4-Air at 75.7%. On MATH, GLM-4-Air leads at 57.9% vs LLaMA 3 (8B/70B) at 50.4%. On MMLU, LLaMA 3 (8B/70B) leads at 82% vs GLM-4-Air at 81.9%. On GSM8K, LLaMA 3 (8B/70B) leads at 93% vs GLM-4-Air at 90.9%.