Compare AI models

Specifications
Parameters
16B/236B1.5B/7B/72B
Context window
128k4k
Benchmarks
MATH
75.7%85.9%
GSM8K
94.9%95.9%
Overview
CompanyDeepSeekQwen
Release dateJun 14 2024Sep 19 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Qwen2.5-Math leads DeepSeek Coder V2 on 2 of the 2 benchmarks they both report (MATH, GSM8K). DeepSeek Coder V2 shipped 97 days before Qwen2.5-Math, so benchmark comparisons should account for the intervening progress.

DeepSeek Coder V2 has 16B/236B parameters, while Qwen2.5-Math has 1.5B/7B/72B. Context windows are 128k (DeepSeek Coder V2) vs 4k (Qwen2.5-Math).

On MATH, Qwen2.5-Math leads at 85.9% vs DeepSeek Coder V2 at 75.7%. On GSM8K, Qwen2.5-Math leads at 95.9% vs DeepSeek Coder V2 at 94.9%.