Compare AI models

Specifications
Parameters
671B—
Context window
128k128k
API pricing
Cheapest input
—$0.40Mistral
Cheapest output
—$2.00Mistral
—$180Mistral · zdr
Benchmarks
LiveCodeBench
65.9%30.3%
MATH-500
97.3%91%
GPQA Diamond
71.5%57.1%
MMLU-Pro
84%77.2%
Overview
CompanyDeepSeekMistral
Release dateJan 20 2025May 7 2025
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

DeepSeek R1 leads Mistral Medium 3 on 4 of the 4 benchmarks they both report (LiveCodeBench, MATH-500, GPQA Diamond, MMLU-Pro). DeepSeek R1 shipped 107 days before Mistral Medium 3, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (DeepSeek R1) vs 128k (Mistral Medium 3). DeepSeek R1 is open weight, while Mistral Medium 3 is closed.

On LiveCodeBench, DeepSeek R1 leads at 65.9% vs Mistral Medium 3 at 30.3%. On MATH-500, DeepSeek R1 leads at 97.3% vs Mistral Medium 3 at 91%. On GPQA Diamond, DeepSeek R1 leads at 71.5% vs Mistral Medium 3 at 57.1%. On MMLU-Pro, DeepSeek R1 leads at 84% vs Mistral Medium 3 at 77.2%.