AGIEval English Leaderboard

AGIEval English — human-level reasoning tasks from standardized exams like SAT, LSAT, and civil service exams.

Data from LayerLens

As of November 21, 2026, the top-scoring model on AGIEval English is Gemini 3.1 Pro Preview at 94.0%, followed by Claude Opus 5 at 93.5% and Gemini 3.7 Flash at 93.4%. 137 models have been evaluated on this benchmark.

Last updated: November 21, 2026

8 Ways to Use Fewer Tokens

Models

137

Best Score

94.0

Average

81.2

Std Dev

12.6

Categories
Reasoning and Logic
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
AGIEval English
Actions
$1.000
$6.000
94.0
$5.000
$25.000
93.5
$0.750
$3.750
93.4
$1.500
$9.000
93.0
$2.000
$6.000
92.9
$5.000
$30.000
92.8
$1.475
$4.425
92.8
$0.500
$3.000
92.7
$2.000
$6.000
92.7
$0.750
$3.750
92.6
$0.500
$3.000
92.5
$2.000
$6.000
92.1
$0.007
$0.950
92.1
$0.320
$1.280
92.0
$0.000
$0.000
91.8
$0.490
$2.500
91.7
$0.490
$2.500
91.7
$0.390
$0.900
91.4
$0.040
$0.078
91.4
$1.250
$10.000
91.4
$0.600
$3.000
91.4
$1.000
$10.000
91.1
$0.216
$0.647
91.1
$5.000
$25.000
91.0
$5.000
$25.000
91.0
$0.400
$1.760
91.0
$2.000
$8.000
90.9
$0.195
$0.900
90.6
$0.899
$2.825
90.5
$1.200
$4.000
90.4
$0.260
$0.900
90.3
$1.000
$4.050
90.2
$0.320
$1.400
90.1
$0.320
$1.400
90.1
$0.014
$0.029
90.1
$0.250
$0.900
89.9
$3.000
$15.000
89.5
$0.065
$0.260
89.5
$0.065
$0.750
89.5
$0.600
$1.920
89.4
$0.600
$1.920
89.4
$2.000
$10.000
89.3
$0.450
$1.200
89.2
$0.209
$0.310
89.0
$0.209
$0.310
89.0
$0.500
$2.150
89.0
$0.270
$0.410
89.0
$5.000
$25.000
89.0
$0.300
$0.900
88.8
$0.050
$0.700
88.7
$0.430
$1.750
88.6
$0.430
$1.750
88.6
$0.188
$1.125
88.6
$0.030
$0.130
88.6
$10.000
$50.000
88.5
$0.450
$0.900
88.4
$0.900
$0.900
88.2
$0.900
$0.900
88.2
$0.230
$0.960
88.2
$5.000
$25.000
88.0
$0.000
$0.000
87.9
$0.600
$2.500
87.9
$0.100
$0.100
87.7
$0.125
$1.000
87.1
$0.200
$1.200
87.1
$0.200
$0.900
86.1
$3.000
$15.000
86.0
$3.000
$15.000
85.7
$0.270
$0.410
85.7
$0.020
$0.096
85.7
$0.600
$2.200
85.0
$0.270
$1.000
84.7
$0.270
$1.000
84.7
$0.090
$0.900
84.4
$0.064
$0.224
84.4
$0.064
$0.224
84.4
$0.300
$2.500
84.3
$0.030
$0.100
83.9
$0.270
$0.950
83.4
$0.550
$2.200
83.3
$0.120
$0.500
83.1
$0.120
$0.500
83.1
$0.300
$0.900
82.6
$0.050
$0.190
82.5
$0.000
$0.000
81.5
$0.650
$3.250
81.1
$0.250
$0.950
80.8
$0.250
$0.950
80.8
$0.000
$0.000
80.8
$0.200
$0.880
80.7
$0.600
$2.500
80.7
$0.060
$0.240
80.6
$0.060
$0.240
80.6
$0.130
$0.850
80.5
$0.017
$0.085
80.0
$1.000
$5.000
79.7
$0.780
$3.900
79.1
$0.025
$0.200
78.0
$0.260
$1.560
77.8
$1.000
$5.000
76.9
$0.250
$1.500
76.8
$0.200
$0.600
76.7
$0.014
$0.028
76.2
$1.250
$7.500
76.2
$0.300
$2.500
74.2
$0.300
$2.500
74.2
$0.500
$1.500
74.0
$0.440
$2.200
71.9
$1.750
$14.000
71.7
$1.500
$7.500
71.5
$0.440
$2.200
70.3
$2.500
$10.000
70.1
$2.000
$8.000
70.0
$0.625
$5.000
69.4
$2.500
$10.000
69.4
$0.070
$0.280
69.1
$0.288
$0.320
68.3
$0.070
$0.140
68.2
$0.000
$0.000
67.8
$0.075
$0.400
67.0
$0.060
$0.240
65.8
$0.800
$3.200
65.5
$0.080
$0.160
65.1
$1.000
$3.000
64.3
$1.000
$3.000
64.3
$0.375
$2.250
63.0
$0.660
$0.800
63.0
$0.100
$0.280
62.5
$0.075
$0.200
62.3
$0.100
$0.400
58.0
$0.100
$0.400
58.0
$0.035
$0.140
55.6
$0.300
$0.900
48.3
$0.020
$0.040
46.9
$0.100
$0.625
46.5
$0.100
$0.300
27.4
$0.030
$0.050
26.5

Pricing from OpenRouter. Benchmarks from Artificial Analysis.

About AGIEval English

AGIEval English — human-level reasoning tasks from standardized exams like SAT, LSAT, and civil service exams.

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

Frequently Asked Questions

AGIEval English — human-level reasoning tasks from standardized exams like SAT, LSAT, and civil service exams.
As of November 21, 2026, Gemini 3.1 Pro Preview leads the AGIEval English leaderboard with a score of 94.0. Rankings change as new models are released and evaluated.
Currently 137 models have been evaluated on AGIEval English, with an average score of 81.2 and standard deviation of 12.6.
Benchmark scores are updated when new evaluations are published by our data sources (Artificial Analysis and LayerLens). Pricing data is refreshed daily from OpenRouter.