Compare AI models

Specifications
Parameters
—6B
Context window
128k32k
Benchmarks
HumanEval
88.4%9.8%
MATH
76.1%6.9%
MMLU
87.5%45.2%
Overview
CompanySpaceXSIZ.ai
Release dateAug 13 2024Jun 25 2023
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Grok‑2 leads ChatGLM2-6B on 3 of the 3 benchmarks they both report (HumanEval, MATH, MMLU). ChatGLM2-6B shipped 415 days before Grok‑2, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Grok‑2) vs 32k (ChatGLM2-6B). Grok‑2 is closed, while ChatGLM2-6B is open weight.

On HumanEval, Grok‑2 leads at 88.4% vs ChatGLM2-6B at 9.8%. On MATH, Grok‑2 leads at 76.1% vs ChatGLM2-6B at 6.9%. On MMLU, Grok‑2 leads at 87.5% vs ChatGLM2-6B at 45.2%.