Compare AI models

Specifications
Parameters
7B7B
Context window
4k32k
Benchmarks
MATH
51.7%13.2%
Overview
CompanyDeepSeekMistral
Release dateFeb 5 2024May 22 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek-Math leads Mistral 7B v0.3 on 1 of the 1 benchmark they both report (MATH). DeepSeek-Math shipped 107 days before Mistral 7B v0.3, so benchmark comparisons should account for the intervening progress.

DeepSeek-Math has 7B parameters, while Mistral 7B v0.3 has 7B. Context windows are 4k (DeepSeek-Math) vs 32k (Mistral 7B v0.3).

On MATH, DeepSeek-Math leads at 51.7% vs Mistral 7B v0.3 at 13.2%.