Compare AI models

Specifications
Parameters
7B/67B6B
Context window
4k8k
Benchmarks
HumanEval
73.8%58.5%
MATH
32.6%25.7%
MMLU
71.1%61.4%
GSM8K
84.1%72.3%
Overview
CompanyDeepSeekZ.ai
Release dateNov 29 2023Oct 27 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek-LLM leads ChatGLM3-6B on 4 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). ChatGLM3-6B shipped 33 days before DeepSeek-LLM, so benchmark comparisons should account for the intervening progress.

DeepSeek-LLM has 7B/67B parameters, while ChatGLM3-6B has 6B. Context windows are 4k (DeepSeek-LLM) vs 8k (ChatGLM3-6B).

On HumanEval, DeepSeek-LLM leads at 73.8% vs ChatGLM3-6B at 58.5%. On MATH, DeepSeek-LLM leads at 32.6% vs ChatGLM3-6B at 25.7%. On MMLU, DeepSeek-LLM leads at 71.1% vs ChatGLM3-6B at 61.4%. On GSM8K, DeepSeek-LLM leads at 84.1% vs ChatGLM3-6B at 72.3%.