Compare AI models

Specifications
Parameters
16B14B
Context window
32k2k
Benchmarks
HumanEval
57.3%32.3%
MATH
27.9%24.8%
MMLU
55.7%66.3%
GSM8K
72%61.3%
Overview
CompanyDeepSeekQwen
Release dateMay 16 2024Sep 25 2023
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek V2 Lite leads Qwen-14B on 3 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). Qwen-14B shipped 234 days before DeepSeek V2 Lite, so benchmark comparisons should account for the intervening progress.

DeepSeek V2 Lite has 16B parameters, while Qwen-14B has 14B. Context windows are 32k (DeepSeek V2 Lite) vs 2k (Qwen-14B).

On HumanEval, DeepSeek V2 Lite leads at 57.3% vs Qwen-14B at 32.3%. On MATH, DeepSeek V2 Lite leads at 27.9% vs Qwen-14B at 24.8%. On MMLU, Qwen-14B leads at 66.3% vs DeepSeek V2 Lite at 55.7%. On GSM8K, DeepSeek V2 Lite leads at 72% vs Qwen-14B at 61.3%.