Skip to content

Benchmark

GSO

Coding benchmarks compare models on programming tasks. Scores are percentages.

Models measured
30
Scores recorded
33
Sources
1
IQ weight
2.99

Score distribution

24.4%
Mean
14.7%
Median
24.7%
Std dev
3.6%
10th percentile
76.5%
90th percentile
0%
Min
0%88.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.99 for chat models, measured over 30 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 Fable 5.1· Epoch AI Benchmarking Hub data88.2%
2GPT-6 Astra· Epoch AI Benchmarking Hub data79.4%
3Claude Fable 5· Epoch AI Benchmarking Hub data78.4%
4GPT-5.6 Sol· Epoch AI Benchmarking Hub data76.5%
5Claude Opus 4.8· Epoch AI Benchmarking Hub data47.1%
6Claude Opus 4.7· Epoch AI Benchmarking Hub data44.1%
7Claude Opus 4.6high· Epoch AI Benchmarking Hub data41.2%
8GPT 5.5xhigh· Epoch AI Benchmarking Hub data40.2%
9Claude Sonnet 5· Epoch AI Benchmarking Hub data37.3%
10Claude Opus 4.6· Epoch AI Benchmarking Hub data33.3%
11GPT 5.4xhigh· Epoch AI Benchmarking Hub data31.4%
12GPT 5.2high· Epoch AI Benchmarking Hub data27.4%
13Claude Opus 4.5· Epoch AI Benchmarking Hub data26.5%
14GPT 5.4high· Epoch AI Benchmarking Hub data25.5%
15Gemini 3.1 Pro Preview· Epoch AI Benchmarking Hub data22.6%
16Gemini 3 Pro Preview· Epoch AI Benchmarking Hub data18.6%
17Claude Sonnet 4.5· Epoch AI Benchmarking Hub data14.7%
18GPT 5.1· Epoch AI Benchmarking Hub data13.7%
19GPT 5.1high· Epoch AI Benchmarking Hub data13.7%
20Gemini 3 Flash· Epoch AI Benchmarking Hub data9.8%
21OpenAI o3high· Epoch AI Benchmarking Hub data8.8%
22Claude Opus 4· Epoch AI Benchmarking Hub data6.9%
23GPT 5high· Epoch AI Benchmarking Hub data6.9%
24Claude Sonnet 4· Epoch AI Benchmarking Hub data4.9%
25Kimi K2 Instruct· Epoch AI Benchmarking Hub data4.9%
26Qwen3 Coder· Epoch AI Benchmarking Hub data4.9%
27Claude 3.5 Sonnet· Epoch AI Benchmarking Hub data4.6%
28Gemini 2.5 Pro· Epoch AI Benchmarking Hub data3.9%
29Claude Sonnet 3.7· Epoch AI Benchmarking Hub data3.8%
30o4 Minihigh· Epoch AI Benchmarking Hub data3.6%
31GLM 4.5 Air· Epoch AI Benchmarking Hub data2.9%
32o3 Mini· Epoch AI Benchmarking Hub data1.3%
33GPT 4o· Epoch AI Benchmarking Hub data0%

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