Compare AI models

Specifications
Parameters
—6B
Context window
128k32k
Benchmarks
HumanEval
74.1%9.8%
MATH
50.6%6.9%
MMLU
81.3%45.2%
GSM8K
90%25.9%
Overview
CompanySpaceXSIZ.ai
Release dateMar 28 2024Jun 25 2023
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Grok‑1.5 leads ChatGLM2-6B on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). ChatGLM2-6B shipped 277 days before Grok‑1.5, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Grok‑1.5) vs 32k (ChatGLM2-6B). Grok‑1.5 is closed, while ChatGLM2-6B is open weight.

On HumanEval, Grok‑1.5 leads at 74.1% vs ChatGLM2-6B at 9.8%. On MATH, Grok‑1.5 leads at 50.6% vs ChatGLM2-6B at 6.9%. On MMLU, Grok‑1.5 leads at 81.3% vs ChatGLM2-6B at 45.2%. On GSM8K, Grok‑1.5 leads at 90% vs ChatGLM2-6B at 25.9%.