Compare AI models

Specifications
Parameters
123B—
Context window
256k—
API pricing
Cheapest input
$0.40Mistral—
Cheapest output
$2.00Mistral—
$180Mistral—
Benchmarks
SWE-Bench Multilingual
61.3%56.9%
Terminal-Bench 2.0
32.6%40%
Overview
CompanyMistralSpaceXSI
Release dateDec 9 2025Oct 29 2025
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

Devstral 2 (123B) and Composer are evenly matched across the 2 benchmarks they both report (SWE-Bench Multilingual, Terminal-Bench 2.0). Composer shipped 41 days before Devstral 2 (123B), so benchmark comparisons should account for the intervening progress.

Devstral 2 (123B) is open weight, while Composer is closed.

On SWE-Bench Multilingual, Devstral 2 (123B) leads at 61.3% vs Composer at 56.9%. On Terminal-Bench 2.0, Composer leads at 40% vs Devstral 2 (123B) at 32.6%.