Skip to content

Benchmark

Dubesor Tech

Scores are percentages.

Models measured
214
Scores recorded
255
Sources
1
IQ weight
1.67

Score distribution

55.2%
Mean
61.2%
Median
24.4%
Std dev
16.5%
10th percentile
81.1%
90th percentile
2.3%
Min
2.3%95.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.67 for chat models, measured over 214 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Grok 4.20reasoning· Dubesor LLM Benchmark79.6%
52Kimi K2 Thinking· Dubesor LLM Benchmark79.6%
53o3 Mini· Dubesor LLM Benchmark79.6%
54Qwen3.5 122B-A10B· Dubesor LLM Benchmark79.6%
55Qwen3.5 397B A17B· Dubesor LLM Benchmark79.6%
56Qwen3.5 Plus· Dubesor LLM Benchmark79.6%
57GPT 4· Dubesor LLM Benchmark76.8%
58GLM 4.7thinking· Dubesor LLM Benchmark75.9%
59GPT-5 Chat· Dubesor LLM Benchmark75.9%
60DeepSeek V2.5· Dubesor LLM Benchmark75.4%
61Gemma 4 26B A4B· Dubesor LLM Benchmark75.4%
62GLM 4.6thinking· Dubesor LLM Benchmark75.4%
63LongCat Flash Chat· Dubesor LLM Benchmark75.4%
64Qwen3 Coder Plus· Dubesor LLM Benchmark75.4%
65GPT 4.1· Dubesor LLM Benchmark75.3%
66MiMo V2 Flash· Dubesor LLM Benchmark75.3%
67gemini-2.5-flash-preview-04-17thinking· Dubesor LLM Benchmark75%
68GLM 5· Dubesor LLM Benchmark75%
69GLM 5thinking· Dubesor LLM Benchmark75%
70DeepSeek V3.1· Dubesor LLM Benchmark74.3%
71DeepSeek V3.2· Dubesor LLM Benchmark74.3%
72GLM 4.6vthinking· Dubesor LLM Benchmark74.3%
73Grok 3 Mini· Dubesor LLM Benchmark74.3%
74Grok Code Fast 1· Dubesor LLM Benchmark74.3%
75Mistral Large 3· Dubesor LLM Benchmark74.3%
76Llama 3.3 Nemotron Super 49B v1.5thinking· Dubesor LLM Benchmark71.8%
77DeepSeek V3.1thinking· Dubesor LLM Benchmark71.7%
78Claude Haiku 3.5· Dubesor LLM Benchmark71.6%
79Claude Sonnet 3.7· Dubesor LLM Benchmark71.4%
80Devstral 2 123b Instruct 2512· Dubesor LLM Benchmark71.2%
81Qwen 3 235B A22B Thinking 2507· Dubesor LLM Benchmark71.2%
82Step 3.5 Flash· Dubesor LLM Benchmark71.2%
83Qwen3 235B A22B Instruct 2507· Dubesor LLM Benchmark71.1%
84Grok 4 Fastreasoning· Dubesor LLM Benchmark70.8%
85Qwen 3 235B A22B· Dubesor LLM Benchmark70.8%
86Grok 3 Minihigh· Dubesor LLM Benchmark70.7%
87o1· Dubesor LLM Benchmark70.2%
88GPT 5.1· Dubesor LLM Benchmark70.1%
89GPT-5.1 Codex· Dubesor LLM Benchmark70.1%
90Kimi K2 Instruct· Dubesor LLM Benchmark70.1%
91Mistral Large 2· Dubesor LLM Benchmark70.1%
92r1-1776· Dubesor LLM Benchmark70.1%
93OpenAI o3· Dubesor LLM Benchmark69.9%
94Kimi K2.5· Dubesor LLM Benchmark69.6%
95gemini-1.5-pro-002· Dubesor LLM Benchmark69.5%
96Qwen3 32Bthinking· Dubesor LLM Benchmark67.6%
97Llama 3.1 405B Instruct· Dubesor LLM Benchmark67.4%
98GPT-4.1 mini· Dubesor LLM Benchmark67%
99GPT-5 nano· Dubesor LLM Benchmark66.8%
100Qwen2.5 72B Instruct· Dubesor LLM Benchmark66.8%

Showing 51–100 of 255 results.