Compare AI models

Specifications
Parameters
—6B
Context window
128k32k
Benchmarks
HumanEval
83.7%9.8%
MATH
64.3%6.9%
MMLU
84.7%45.2%
Overview
CompanyOpenAIZ.ai
Release dateNov 6 2023Jun 25 2023
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

GPT-4 Turbo leads ChatGLM2-6B on 3 of the 3 benchmarks they both report (HumanEval, MATH, MMLU). ChatGLM2-6B shipped 134 days before GPT-4 Turbo, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (GPT-4 Turbo) vs 32k (ChatGLM2-6B). GPT-4 Turbo is closed, while ChatGLM2-6B is open weight.

On HumanEval, GPT-4 Turbo leads at 83.7% vs ChatGLM2-6B at 9.8%. On MATH, GPT-4 Turbo leads at 64.3% vs ChatGLM2-6B at 6.9%. On MMLU, GPT-4 Turbo leads at 84.7% vs ChatGLM2-6B at 45.2%.