Skip to content

Benchmark

Dubesor Utility

Scores are percentages.

Models measured
224
Scores recorded
266
Sources
1
IQ weight
1.16

Score distribution

53.7%
Mean
55%
Median
17.9%
Std dev
28.7%
10th percentile
76.8%
90th percentile
4%
Min
4%87.4%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.16 for chat models, measured over 224 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Grok 4.20reasoning· Dubesor LLM Benchmark87.4%
2Kimi K2.5thinking· Dubesor LLM Benchmark87.4%
3Claude Opus 4.7· Dubesor LLM Benchmark84.5%
4GPT-5 Chat· Dubesor LLM Benchmark83.6%
5DeepSeek V3.2 Speciale· Dubesor LLM Benchmark82.6%
6GPT 4o· Dubesor LLM Benchmark82.6%
7Grok 4· Dubesor LLM Benchmark82.6%
8Kimi K2 Thinking· Dubesor LLM Benchmark82.6%
9Gemini 2.5 Flash Preview 05-20 (no think)thinking· Dubesor LLM Benchmark82.2%
10Llama 3.1 405B Instruct· Dubesor LLM Benchmark81.5%
11Claude Sonnet 4.6· Dubesor LLM Benchmark80.1%
12Gemini 3.1 Flash Lite· Dubesor LLM Benchmark79.7%
13GPT-4 Turbo· Dubesor LLM Benchmark79.5%
14Gemini 2.5 Pro· Dubesor LLM Benchmark79%
15Gemini 3 Flash· Dubesor LLM Benchmark78.8%
16Gemini 3 Flashthinking· Dubesor LLM Benchmark78.8%
17DeepSeek V3.1· Dubesor LLM Benchmark78.6%
18Gemini 2.0 Flash· Dubesor LLM Benchmark78.1%
19gemini-2.5-flash-preview-04-17· Dubesor LLM Benchmark77.8%
20gemini-2.5-flash-preview-04-17thinking· Dubesor LLM Benchmark77.8%
21Gemma 4 31B IT· Dubesor LLM Benchmark77.7%
22GLM 4.5thinking· Dubesor LLM Benchmark77.7%
23ChatGPT-4o Latest· Dubesor LLM Benchmark77.4%
24Claude Opus 4.5· Dubesor LLM Benchmark77.4%
25DeepSeek R1· Dubesor LLM Benchmark77.4%
26Gemini 2.5 Pro Preview 05-06· Dubesor LLM Benchmark77.4%
27Kimi K2.5· Dubesor LLM Benchmark77.4%
28Gemma 2 27B IT· Dubesor LLM Benchmark76.8%
29DeepSeek V3.2· Dubesor LLM Benchmark76.4%
30Kimi K2 Instruct· Dubesor LLM Benchmark76.4%
31GPT-4.5 Preview· Dubesor LLM Benchmark76.3%
32GPT 4.1· Dubesor LLM Benchmark75.6%
33Gemini 1.5 Pro· Dubesor LLM Benchmark74.1%
34Qwen 3 235B A22B Thinking 2507· Dubesor LLM Benchmark72.7%
35Grok 3 Minihigh· Dubesor LLM Benchmark72.6%
36Mistral Large 3· Dubesor LLM Benchmark72.6%
37GLM 4.6thinking· Dubesor LLM Benchmark72.4%
38Mistral Large 2· Dubesor LLM Benchmark72.3%
39GLM 4.5 Air· Dubesor LLM Benchmark72%
40GPT-4.1 mini· Dubesor LLM Benchmark72%
41gemini-1.5-flash-002· Dubesor LLM Benchmark71.9%
42Gemma 3 27B IT· Dubesor LLM Benchmark71.9%
43Llama 3.3 Nemotron Super 49B v1thinking· Dubesor LLM Benchmark71.9%
44GPT 5· Dubesor LLM Benchmark71.7%
45DeepSeek V3· Dubesor LLM Benchmark71.6%
46Llama 3.3 70B Instruct· Dubesor LLM Benchmark71.6%
47Gemini 2.5 Flash-Lite· Dubesor LLM Benchmark71.4%
48GLM 4.5· Dubesor LLM Benchmark71.3%
49GLM 4.6· Dubesor LLM Benchmark71.3%
50Claude Sonnet 3.7· Dubesor LLM Benchmark70.9%

Showing 1–50 of 266 results.