Compare AI models

Specifications
Parameters
—1T
Context window
128k256k
API pricing
Cheapest input
—$0.49AtlasCloud
Cheapest output
—$2.50AtlasCloud
—$223AtlasCloud · int4
Benchmarks
GPQA Diamond
65.7%87.6%
Overview
CompanyMistralMoonshot AI
Release dateAug 12 2025Jan 27 2026
AccessClosedOpen Weight
Model detailsView modelView model

Frequently asked questions

Kimi K2.5 leads Mistral Medium 3.1 on 1 of the 1 benchmark they both report (GPQA Diamond). Mistral Medium 3.1 shipped 168 days before Kimi K2.5, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (Mistral Medium 3.1) vs 256k (Kimi K2.5). Mistral Medium 3.1 is closed, while Kimi K2.5 is open weight.

On GPQA Diamond, Kimi K2.5 leads at 87.6% vs Mistral Medium 3.1 at 65.7%.