Compare AI models

Specifications
Parameters
16B/236B6B
Context window
128k8k
Benchmarks
HumanEval
90.2%58.5%
MATH
75.7%25.7%
MMLU
79.2%61.4%
GSM8K
94.9%72.3%
Overview
CompanyDeepSeekZ.ai
Release dateJun 14 2024Oct 27 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek Coder V2 leads ChatGLM3-6B on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). ChatGLM3-6B shipped 231 days before DeepSeek Coder V2, so benchmark comparisons should account for the intervening progress.

DeepSeek Coder V2 has 16B/236B parameters, while ChatGLM3-6B has 6B. Context windows are 128k (DeepSeek Coder V2) vs 8k (ChatGLM3-6B).

On HumanEval, DeepSeek Coder V2 leads at 90.2% vs ChatGLM3-6B at 58.5%. On MATH, DeepSeek Coder V2 leads at 75.7% vs ChatGLM3-6B at 25.7%. On MMLU, DeepSeek Coder V2 leads at 79.2% vs ChatGLM3-6B at 61.4%. On GSM8K, DeepSeek Coder V2 leads at 94.9% vs ChatGLM3-6B at 72.3%.