Mistral Large 4 Preview

Released2026-10-06
Price per 1MNot listed
Overall score57.1 · rank 58
ProviderMistral
WeightsOpen weights

Scores

DimensionScoreRankBenchmarks
Overall: Balanced 57.1 58 21
Overall: Coding 59.5 35 16
Overall: Reasoning 55.0 73 11
Overall: Agentic work 58.6 48 15
Overall: Frontend 60.0 21 6
Coding 60.2 26 5
Agents 58.1 52 4
Reasoning 57.7 66 6
Math 57.3 53 4
Knowledge 45.1 115 1
Frontend 59.9 22 1

Configurations

ConfigurationBalancedBenchmarks
Unspecified setting 59.8 ±2.4 10
High 56.3 ±1.9 11

Evidence

Coding 60.2 ±3.2

BenchmarkResultSourceDate
FrontierCode 39.4% Epoch AI 2026-11-11
LiveBench 2.0 Coding 79.4% LiveBench 2026-11-06
WeirdML v3 4.5% Epoch AI 2026-11-11
MirrorCode 27.8% Epoch AI 2026-10-21
SciCode 54.2% Epoch AI 2026-11-11

Agents 58.1 ±1.6

BenchmarkResultSourceDate
APEX-Agents 51.6% Epoch AI 2026-11-11
LiveBench 2.0 Agentic 24.7% LiveBench 2026-11-06
LiveBench 2.0 Browser Use 17.4% LiveBench 2026-11-06
LiveBench 2.0 Research 80.3% LiveBench 2026-11-06

Reasoning 57.7 ±3.0

BenchmarkResultSourceDate
GPQA Diamond 88.1% Epoch AI 2026-10-12
CritPt 10.6% Epoch AI 2026-11-11
Chess Puzzles 33.0% Epoch AI 2026-10-15
LiveBench 2.0 Reasoning 85.2% LiveBench 2026-11-06
Mystery Game Puzzles 17.0% Epoch AI 2026-10-13
LiveBench 2.0 Data Analysis 76.6% LiveBench 2026-11-06

Math 57.3 ±2.6

BenchmarkResultSourceDate
FrontierMath Tiers 1-3 58.6% Epoch AI 2026-10-12
FrontierMath Tier 4 26.8% Epoch AI 2026-10-12
ProofBench 10.0% Epoch AI 2026-11-11
OTIS Mock AIME 2024-2025 98.3% Epoch AI 2026-10-12

Knowledge 45.1 ±3.0

BenchmarkResultSourceDate
SimpleQA Verified 20.0% Epoch AI 2026-10-09

Frontend 59.9 ±2.7

BenchmarkResultSourceDate
LiveBench 2.0 Design 76.3% LiveBench 2026-11-06

Names matched to this model

mistral-large-4-0, mistral-large-4-preview

Compare Mistral Large 4 Preview with other models