Compare AI models

Specifications
Parameters
16B—
Context window
32k128k
API pricing
Cheapest input
—$0.40Mistral
Cheapest output
—$2.00Mistral
—$180Mistral · zdr
Benchmarks
HumanEval
57.3%92.1%
Overview
CompanyDeepSeekMistral
Release dateMay 16 2024May 7 2025
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Mistral Medium 3 leads DeepSeek V2 Lite on 1 of the 1 benchmark they both report (HumanEval). DeepSeek V2 Lite shipped 356 days before Mistral Medium 3, so benchmark comparisons should account for the intervening progress.

Context windows are 32k (DeepSeek V2 Lite) vs 128k (Mistral Medium 3). DeepSeek V2 Lite is open weight, while Mistral Medium 3 is closed.

On HumanEval, Mistral Medium 3 leads at 92.1% vs DeepSeek V2 Lite at 57.3%.