Note: Overall leaderboard rankings may not reflect true model quality — individual benchmarks give a clearer picture. ARC-Challenge MMLU GPQA GSM8K Artificial Analysis Intelligence Index v4.1
← Back to leaderboard

Mistral AI: devstral-2

1 benchmarks
CompanyMistral AI
Country of originFrance

Leaderboard position

#396
Overall rank (of 1126 models)
1328
Elo rating
1
Benchmarks
Top 35.2%
Percentile
In Mistral AI (of 24 models)

Benchmark highlights

Nearest competitor from another company

The model from another company with the most similar benchmark scores.

appeared as nearest in 1 of 1 benchmarks

1/1 100.0%

Benchmarks

Predicted scores

Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.

Benchmark Predicted Confidence Based on
ARC-Challenge ~84.6 50% —
ARC-Easy ~94.8 50% —
HellaSwag ~95.4 50% —
PIQA ~84.39 50% —
WinoGrande ~82.88 50% —
Humanity's Last Exam (no tools) ~20.01 50% —
Artificial Analysis Intelligence Index v5.0.1 ~12 50% —
MATH-500 ~95.69 50% —
AIME 2025 ~82.08 50% —
AIME 2026 ~82.51 50% —
HMMT Feb 2026 ~74.1 50% —
GPQA Diamond ~68.55 50% —
MMLU-Pro ~73.41 50% —
IFEval ~89.45 50% —
IFBench ~61.22 50% —
HumanEval ~95.9 50% —
MBPP ~93.7 50% —
Terminal-Bench v2.1 ~37.56 50% —
MMLU ~80.94 50% —
SWE-bench Verified ~54.16 50% —
SWE-bench Pro ~32.05 50% —
SciCode ~34.25 50% —
AA-LCR ~52.25 50% —
Humanity's Last Exam (with tools) ~27.17 50% —
GSM8K ~99.6 50% —