Skip to content

Benchmark

Dubesor STEM

Scores are percentages.

Models measured
225
Scores recorded
268
Sources
1
IQ weight
1.64

Score distribution

51.4%
Mean
55.8%
Median
23.1%
Std dev
15.8%
10th percentile
79.9%
90th percentile
0.5%
Min
0.5%93.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.64 for chat models, measured over 225 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1o1· Dubesor LLM Benchmark93.8%
2Claude Sonnet 4· Dubesor LLM Benchmark93.2%
3DeepSeek R1· Dubesor LLM Benchmark92.9%
4Claude Opus 4.1· Dubesor LLM Benchmark90.9%
5Claude Opus 4.1thinking· Dubesor LLM Benchmark89.6%
6GPT 5· Dubesor LLM Benchmark87.9%
7Claude Opus 4thinking· Dubesor LLM Benchmark87.2%
8DeepSeek V3.2 Speciale· Dubesor LLM Benchmark87.1%
9Gemini 3.1 Pro Preview· Dubesor LLM Benchmark86.6%
10Claude Opus 4.7· Dubesor LLM Benchmark86.1%
11Claude Sonnet 3.7thinking· Dubesor LLM Benchmark85.8%
12Gemini 3 Pro Preview· Dubesor LLM Benchmark85.6%
13GPT 5.1· Dubesor LLM Benchmark83.9%
14Claude Opus 4.5· Dubesor LLM Benchmark83.8%
15OpenAI o3· Dubesor LLM Benchmark83.8%
16GPT 5.4· Dubesor LLM Benchmark83.5%
17GPT-4 Turbo· Dubesor LLM Benchmark83.3%
18Qwen 3 235B A22B· Dubesor LLM Benchmark83.3%
19Claude Sonnet 4.6· Dubesor LLM Benchmark83.2%
20Claude Opus 4· Dubesor LLM Benchmark83.1%
21Claude Sonnet 3.7· Dubesor LLM Benchmark83.1%
22Qwen3 235B A22B Instruct 2507· Dubesor LLM Benchmark81.9%
23DeepSeek R1 Zero· Dubesor LLM Benchmark80.9%
24DeepSeek V3.2· Dubesor LLM Benchmark80.8%
25GPT-4.5 Preview· Dubesor LLM Benchmark80.4%
26Grok 3 Minihigh· Dubesor LLM Benchmark80.1%
27GPT 5.2· Dubesor LLM Benchmark80%
28Claude Sonnet 4thinking· Dubesor LLM Benchmark79.9%
29GPT 5.4thinking· Dubesor LLM Benchmark79.9%
30Grok 3· Dubesor LLM Benchmark79.9%
31Claude 3.5 Sonnet· Dubesor LLM Benchmark78.2%
32o4 Mini· Dubesor LLM Benchmark77.7%
33GPT-5.1 Chat· Dubesor LLM Benchmark77.6%
34Claude Opus 4.6· Dubesor LLM Benchmark77.3%
35GPT-5 mini· Dubesor LLM Benchmark77.3%
36MiMo V2 Pro· Dubesor LLM Benchmark77.3%
37Gemma 4 31B IT· Dubesor LLM Benchmark77.2%
38GLM 4.7thinking· Dubesor LLM Benchmark77.2%
39Grok 4.20reasoning· Dubesor LLM Benchmark76.9%
40Llama 3.3 Nemotron Super 49B v1.5thinking· Dubesor LLM Benchmark76.8%
41Claude 3 Opus· Dubesor LLM Benchmark76.5%
42Gemini 2.5 Pro Preview 03-25· Dubesor LLM Benchmark75.7%
43Gemini 2.5 Pro Preview 05-06· Dubesor LLM Benchmark75.7%
44Grok 4· Dubesor LLM Benchmark75.2%
45GLM 4.6thinking· Dubesor LLM Benchmark75.1%
46Qwen3 Max· Dubesor LLM Benchmark74.5%
47Claude Sonnet 4.5· Dubesor LLM Benchmark73.9%
48o3 Mini· Dubesor LLM Benchmark73.7%
49GPT-5.2 Chat· Dubesor LLM Benchmark73.5%
50Kimi K2.5· Dubesor LLM Benchmark73.5%

Showing 1–50 of 268 results.