Compare AI models

Specifications
Context window
128k128k
Benchmarks
HumanEval
88.4%75.7%
MATH
76.1%57.9%
MMLU
87.5%81.9%
Overview
CompanySpaceXSIZ.ai
Release dateAug 13 2024Jun 5 2024
AccessClosedClosed
Model detailsView modelView model

Frequently asked questions

Grok‑2 leads GLM-4-Air on 3 of the 3 benchmarks they both report (HumanEval, MATH, MMLU). GLM-4-Air shipped 69 days before Grok‑2, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Grok‑2) vs 128k (GLM-4-Air).

On HumanEval, Grok‑2 leads at 88.4% vs GLM-4-Air at 75.7%. On MATH, Grok‑2 leads at 76.1% vs GLM-4-Air at 57.9%. On MMLU, Grok‑2 leads at 87.5% vs GLM-4-Air at 81.9%.