Compare AI models

Specifications
Parameters
16B8B
Context window
4k8k
Benchmarks
HumanEval
45.7%36.22%
MMLU
47.2%69.51%
GSM8K
62.2%58.45%
Overview
CompanyDeepSeekNVIDIA
Release dateJan 11 2024Aug 21 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

DeepSeek-MoE leads Mistral NeMo Minitron 8B on 2 of the 3 benchmarks they both report (HumanEval, MMLU, GSM8K). DeepSeek-MoE shipped 223 days before Mistral NeMo Minitron 8B, so benchmark comparisons should account for the intervening progress.

DeepSeek-MoE has 16B parameters, while Mistral NeMo Minitron 8B has 8B. Context windows are 4k (DeepSeek-MoE) vs 8k (Mistral NeMo Minitron 8B).

On HumanEval, DeepSeek-MoE leads at 45.7% vs Mistral NeMo Minitron 8B at 36.22%. On MMLU, Mistral NeMo Minitron 8B leads at 69.51% vs DeepSeek-MoE at 47.2%. On GSM8K, DeepSeek-MoE leads at 62.2% vs Mistral NeMo Minitron 8B at 58.45%.