Mistral Large (latest)

Mistral AI · mistral/mistral-large-latest

ProviderMistral AI
Familymistral-large
Typevlm
Statusactive
Released2024-11-01
Updated2025-12-02

Reported benchmark scores

These scores carry one collection date for the whole card (2026-04) and a source list rather than a source per score, so they cannot be attributed individually. They are shown as reported and are not verified evidence. Benchmark pages carry the reviewed, dated evidence where it exists.
BenchmarkCatalogue standingScore
Aider Polyglot Benchmarkunassessed52.8
AlpacaEvalunassessed38.5
Arena Elo — Codingunassessed1270.0
Arena Elo — Mathunassessed1260.0
Arena Elo — Overall (Text)unassessed1280.0
Artificial Analysis Intelligence Indexunassessed78.0
Artificial Analysis Output Speedunassessed60.0
BIG-Bench Hardunassessed58.9
BBQ (Bias Benchmark for QA)unassessed77.5
ChartQAunassessed74.5
DocVQAunassessed83.5
GPQA Diamondunassessed55.5
GSM8Kunassessed93.0
HELM Safetyunassessed82.8
HumanEvalunassessed84.2
IFEvalunassessed82.0
LiveCodeBenchunassessed34.4
MATH-500unassessed78.5
MathVistaunassessed50.2
MMLU: Astronomyunassessed73.5
MMLU: Biology (subcategory)unassessed81.5
MMLU: Business Ethicsunassessed74.8
MMLU: Chemistry (subcategory)unassessed74.8
MMLU: Clinical Knowledgeunassessed80.5
MMLU: Computer Science (subcategory)unassessed82.2
MMLU: Jurisprudenceunassessed71.2
MMLU: Physics (unresolved key)unassessed78.2
MMLU-Prounassessed69.8
MMLU: Professional Accountingunassessed65.8
MMLU: Professional Lawunassessed70.5
MMMUunassessed56.8
MT-Benchunassessed8.5
MultiPL-Eunassessed75.8
MultiPL-E: C++unassessed75.5
MultiPL-E: C#unassessed78.2
MultiPL-E: Gounassessed72.1
MultiPL-E: Javaunassessed80.2
MultiPL-E: JavaScriptunassessed78.8
MultiPL-E: Juliaunassessed58.8
MultiPL-E: Kotlin (unconfirmed)unassessed70.2
MultiPL-E: Luaunassessed52.8
MultiPL-E: Perlunassessed48.5
MultiPL-E: PHPunassessed75.5
MultiPL-E: Pythonunassessed82.8
MultiPL-E: Runassessed55.2
MultiPL-E: Rubyunassessed62.5
MultiPL-E: Rustunassessed63.5
MultiPL-E: Scalaunassessed57.8
MultiPL-E: Swiftunassessed63.5
MultiPL-E: TypeScriptunassessed77.5
MuSRunassessed26.3
SWE-bench Verifiedunassessed32.5
ToxiGenunassessed86.2
WildBenchunassessed65.8

Data

This card as JSON · Edit on GitHub