Compare AI models

Specifications
Parameters
—8B/49B
Context window
128k128k
These models have no shared benchmark scores.
Overview
CompanyMistralNVIDIA
Release dateJul 10 2025Mar 18 2025
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Devstral Medium and Llama Nemotron (Nano/Super) don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Llama Nemotron (Nano/Super) shipped 114 days before Devstral Medium, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Devstral Medium) vs 128k (Llama Nemotron (Nano/Super)). Devstral Medium is closed, while Llama Nemotron (Nano/Super) is open weight.

Direct benchmark comparisons are unavailable — Devstral Medium and Llama Nemotron (Nano/Super) don't publish scores on any of the same benchmarks.