Skip to content

Benchmark

Gbaeval Procedural score

Scores are percentages.

Models measured
28
Scores recorded
28
Sources
1
IQ weight
1.74

Score distribution

33.4%
Mean
43.5%
Median
27.7%
Std dev
0%
10th percentile
60.8%
90th percentile
0%
Min
0%87.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.74 for chat models, measured over 28 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Opus 5.5· Epoch AI Benchmarking Hub data87.6%
2Claude Fable 5.2· Epoch AI Benchmarking Hub data77.8%
3GPT-6 Astra· Epoch AI Benchmarking Hub data70.1%
4GPT-6.1 Sol· Epoch AI Benchmarking Hub data60.8%
5Gemini 4 Argon· Epoch AI Benchmarking Hub data58.3%
6GPT-5.6 Sol· Epoch AI Benchmarking Hub data55.9%
7Claude Opus 4.8· Epoch AI Benchmarking Hub data53.3%
8Claude Opus 5· Epoch AI Benchmarking Hub data53.3%
9Claude Fable 5· Epoch AI Benchmarking Hub data53.3%
10Claude Opus 4.7· Epoch AI Benchmarking Hub data53.3%
11Claude Sonnet 5· Epoch AI Benchmarking Hub data48.4%
12Claude Sonnet 4.6· Epoch AI Benchmarking Hub data48.4%
13Grok 4.5· Epoch AI Benchmarking Hub data48.4%
14Claude Opus 4.6· Epoch AI Benchmarking Hub data43.5%
15Kimi K3· Epoch AI Benchmarking Hub data39%
16GPT 5.4· Epoch AI Benchmarking Hub data29.3%
17GPT 5.5· Epoch AI Benchmarking Hub data24.4%
18Gemini 3.5 Flash· Epoch AI Benchmarking Hub data19.1%
19Muse Spark 1.1· Epoch AI Benchmarking Hub data9.9%
20Kimi K2.7 Code· Epoch AI Benchmarking Hub data0.00000000000131%
21Grok Build 0.1· Epoch AI Benchmarking Hub data0.00000000000001%
22Gemini 3.1 Pro Preview· Epoch AI Benchmarking Hub data0.00000000000001%
23GLM 5.2· Epoch AI Benchmarking Hub data0.00000000000001%
24Kimi K2.6· Epoch AI Benchmarking Hub data0.00000000000001%
25MiniMax M3· Epoch AI Benchmarking Hub data0.00000000000001%
26Qwen3.7 Max· Epoch AI Benchmarking Hub data0.00000000000001%
27GLM 5.1· Epoch AI Benchmarking Hub data0%
28MiniMax M2.7· Epoch AI Benchmarking Hub data0%

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)