Compare AI models

Specifications
Parameters
7B/67B—
Context window
4k—
Benchmarks
HumanEval
73.8%72%
MATH
32.6%47.9%
MMLU
71.1%81.5%
GSM8K
84.1%87.6%
Overview
CompanyDeepSeekZ.ai
Release dateNov 29 2023Jan 16 2024
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

GLM-4 leads DeepSeek-LLM on 3 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). DeepSeek-LLM shipped 48 days before GLM-4, so benchmark comparisons should account for the intervening progress.

DeepSeek-LLM is open weight, while GLM-4 is closed.

On HumanEval, DeepSeek-LLM leads at 73.8% vs GLM-4 at 72%. On MATH, GLM-4 leads at 47.9% vs DeepSeek-LLM at 32.6%. On MMLU, GLM-4 leads at 81.5% vs DeepSeek-LLM at 71.1%. On GSM8K, GLM-4 leads at 87.6% vs DeepSeek-LLM at 84.1%.