MMLU Leaderboard

Massive Multitask Language Understanding — tests knowledge across 57 subjects.

Data from LayerLens

As of November 21, 2026, the top-scoring model on MMLU is Qwen3.7 Max at 93.7%, followed by GPT-5 at 93.5% and o3 at 93.1%. 69 models have been evaluated on this benchmark.

Last updated: November 21, 2026

8 Ways to Use Fewer Tokens

Models

69

Best Score

93.7

Average

82.8

Std Dev

13.0

Categories
General Knowledge
Sponsored The Grid | Spot Priced LLM API
The Grid | Spot Priced LLM API — Why pay list price when providers will bid for your API usage? Up to 80% off.
View spot prices
Provider
Model
Input $/M
Output $/M
MMLU
Actions
$1.475
$4.425
93.7
$1.250
$10.000
93.5
$2.000
$8.000
93.1
$0.390
$0.900
92.7
$0.260
$1.560
92.5
$0.320
$1.400
91.9
$0.320
$1.400
91.9
$0.260
$0.900
91.8
$0.600
$1.920
91.7
$0.600
$1.920
91.7
$0.195
$0.900
91.5
$0.065
$0.750
90.7
$0.065
$0.260
90.7
$0.500
$2.150
90.5
$0.125
$1.000
90.5
$0.900
$0.900
90.3
$0.075
$0.400
90.2
$0.450
$0.900
90.2
$0.600
$2.200
90.0
$0.100
$0.100
89.0
$1.250
$7.500
89.0
$0.900
$0.900
88.7
$0.900
$0.900
88.7
$0.550
$2.200
88.3
$0.270
$0.950
88.1
$0.780
$3.900
87.6
$0.030
$0.100
87.6
$0.050
$0.190
87.1
$0.200
$0.900
86.8
$0.025
$0.200
86.8
$0.600
$2.500
86.2
$1.750
$14.000
86.1
$0.250
$0.750
86.0
$0.090
$0.900
85.8
$0.200
$0.880
85.8
$0.300
$2.500
85.7
$0.300
$2.500
85.7
$0.200
$0.600
85.5
$0.625
$5.000
85.5
$0.120
$0.500
85.3
$0.120
$0.500
85.3
$0.000
$0.000
85.2
$0.500
$1.500
85.2
$2.000
$8.000
84.6
$2.500
$10.000
84.1
$0.300
$0.900
83.8
$0.060
$0.400
83.7
$0.060
$0.400
83.7
$0.014
$0.028
83.4
$0.130
$0.850
82.9
$0.440
$2.200
82.6
$0.375
$2.250
81.8
$0.288
$0.320
81.2
$0.000
$0.000
79.2
$0.800
$3.200
78.3
$1.000
$3.000
77.7
$1.000
$3.000
77.7
$0.070
$0.280
77.6
$0.070
$0.140
77.6
$0.075
$0.200
76.0
$0.100
$0.280
73.3
$0.060
$0.240
72.7
$0.035
$0.140
68.9
$0.080
$0.200
66.1
$0.080
$0.200
66.1
$0.100
$0.625
65.4
$0.020
$0.040
63.8
$0.100
$0.300
24.6
$0.030
$0.050
15.3

Pricing from OpenRouter. Benchmarks from Artificial Analysis.

About MMLU

Massive Multitask Language Understanding — tests knowledge across 57 subjects.

This leaderboard shows all models with MMLU benchmark scores, ranked from highest to lowest. Pricing data is included to help you compare performance against cost.

Frequently Asked Questions

Massive Multitask Language Understanding — tests knowledge across 57 subjects.
As of November 21, 2026, Qwen3.7 Max leads the MMLU leaderboard with a score of 93.7. Rankings change as new models are released and evaluated.
Currently 69 models have been evaluated on MMLU, with an average score of 82.8 and standard deviation of 13.0.
Benchmark scores are updated when new evaluations are published by our data sources (Artificial Analysis and LayerLens). Pricing data is refreshed daily from OpenRouter.