Compare AI models

Specifications
Parameters
671B12B
Context window
128k128k
API pricing
Cheapest input
$0.256DeepInfra$0.0166Io Net
Cheapest output
$0.712DeepInfra$0.0216Io Net
$103StreamLake$4.40Io Net · fp16
Benchmarks
MMLU
88.5%68%
Overview
CompanyDeepSeekMistral
Release dateDec 26 2024Jul 18 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek V3 leads Mistral NeMo on 1 of the 1 benchmark they both report (MMLU). Mistral NeMo shipped 161 days before DeepSeek V3, so benchmark comparisons should account for the intervening progress.

DeepSeek V3 has 671B parameters, while Mistral NeMo has 12B. Context windows are 128k (DeepSeek V3) vs 128k (Mistral NeMo).

On MMLU, DeepSeek V3 leads at 88.5% vs Mistral NeMo at 68%.