Compare AI models

Specifications
Parameters
16B/236B110B
Context window
128k32k
Benchmarks
HumanEval
90.2%52.4%
MATH
75.7%49.6%
MMLU
79.2%80.4%
GSM8K
94.9%85.4%
Overview
CompanyDeepSeekQwen
Release dateJun 14 2024Apr 25 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek Coder V2 leads Qwen1.5-110B on 3 of the 4 benchmarks they both report (HumanEval, MATH, MMLU, GSM8K). Qwen1.5-110B shipped 50 days before DeepSeek Coder V2, so benchmark comparisons should account for the intervening progress.

DeepSeek Coder V2 has 16B/236B parameters, while Qwen1.5-110B has 110B. Context windows are 128k (DeepSeek Coder V2) vs 32k (Qwen1.5-110B).

On HumanEval, DeepSeek Coder V2 leads at 90.2% vs Qwen1.5-110B at 52.4%. On MATH, DeepSeek Coder V2 leads at 75.7% vs Qwen1.5-110B at 49.6%. On MMLU, Qwen1.5-110B leads at 80.4% vs DeepSeek Coder V2 at 79.2%. On GSM8K, DeepSeek Coder V2 leads at 94.9% vs Qwen1.5-110B at 85.4%.