Compare AI models

Specifications
Parameters
16B6B
Context window
4k8k
Benchmarks
HumanEval
45.7%58.5%
MATH
15.2%25.7%
MMLU
47.2%61.4%
GSM8K
62.2%72.3%
Overview
CompanyDeepSeekZ.ai
Release dateJan 11 2024Oct 27 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

ChatGLM3-6B leads DeepSeek-MoE on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). ChatGLM3-6B shipped 76 days before DeepSeek-MoE, so benchmark comparisons should account for the intervening progress.

DeepSeek-MoE has 16B parameters, while ChatGLM3-6B has 6B. Context windows are 4k (DeepSeek-MoE) vs 8k (ChatGLM3-6B).

On HumanEval, ChatGLM3-6B leads at 58.5% vs DeepSeek-MoE at 45.7%. On MATH, ChatGLM3-6B leads at 25.7% vs DeepSeek-MoE at 15.2%. On MMLU, ChatGLM3-6B leads at 61.4% vs DeepSeek-MoE at 47.2%. On GSM8K, ChatGLM3-6B leads at 72.3% vs DeepSeek-MoE at 62.2%.