Frontier BenchAI model benchmarks · Jul 2026
ModelsCompareLabsMethodology
All models
MA

Mistral Large 3

Mistral AI

675B dense flagship. Cleanest license for EU data-sovereign deployment.

Open-weightApache 2.0256K contextReleased Jan 2026textvisioncode

API price · per 1M tokens

$0.4/$1.2

input / output

≈ $0.60/M blended

Benchmarks

Bar shows this model's score; tick shows the best score across all tracked models for context.

SWE-bench VerifiedReal GitHub issues, human-verified patches. The standard coding-agent benchmark.
68.0/ best 90.1
GPQA DiamondGoogle-proof graduate-level science Q&A. General reasoning proxy.
78.0/ best 94.4
MMLU-ProMulti-task language understanding across 120+ academic subjects.
78.0/ best 93.0
HumanEval+Functional code generation from docstrings.
89.0/ best 96.0
MATH-500Competition-style math problems.
84.0/ best 97.8

Other Mistral AI models

MA

Mistral Small 4

Open-weight

≈$0.15/M

Built by GLM-5.2 · Scores are directional, drawn from public mid-2026 benchmarks.

SWE-bench, GPQA, MMLU, AIME, HLE etc. are trademarks of their respective owners.