Compare AI models

Specifications
Parameters
47B51B
Context window
32k128k
Benchmarks
ARC-Challenge
59.7%69.2%
HellaSwag
84.4%85.58%
WinoGrande
77.2%84.53%
MMLU
70.6%80.2%
GSM8K
58.4%91.43%
Overview
CompanyMistralNVIDIA
Release dateDec 8 2023Sep 23 2024
AccessOpen WeightOpen Weight
Model detailsView modelView model

Frequently asked questions

Llama 3.1 Nemotron 51B leads Mixtral 8×7B on 5 of the 5 benchmarks they both report. Mixtral 8×7B shipped 290 days before Llama 3.1 Nemotron 51B, so benchmark comparisons should account for the intervening progress.

Mixtral 8×7B has 47B parameters, while Llama 3.1 Nemotron 51B has 51B. Context windows are 32k (Mixtral 8×7B) vs 128k (Llama 3.1 Nemotron 51B).

On ARC-Challenge, Llama 3.1 Nemotron 51B leads at 69.2% vs Mixtral 8×7B at 59.7%. On HellaSwag, Llama 3.1 Nemotron 51B leads at 85.58% vs Mixtral 8×7B at 84.4%. On WinoGrande, Llama 3.1 Nemotron 51B leads at 84.53% vs Mixtral 8×7B at 77.2%. On MMLU, Llama 3.1 Nemotron 51B leads at 80.2% vs Mixtral 8×7B at 70.6%. On GSM8K, Llama 3.1 Nemotron 51B leads at 91.43% vs Mixtral 8×7B at 58.4%.