Skip to content

Benchmark

Dubesor Reason

Scores are percentages.

Models measured
218
Scores recorded
261
Sources
1
IQ weight
1.50

Score distribution

44.7%
Mean
42.8%
Median
24.3%
Std dev
11.2%
10th percentile
76.9%
90th percentile
0%
Min
0%93.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.50 for chat models, measured over 218 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201Yi Lightning· Dubesor LLM Benchmark22.7%
202Reka Core· Dubesor LLM Benchmark22.6%
203Gemma 3 27B IT· Dubesor LLM Benchmark22.5%
204Qwen3 14B· Dubesor LLM Benchmark22.5%
205Mistral Small 3.1 24b Instruct 2503· Dubesor LLM Benchmark22.4%
206Gemma 2 27B IT· Dubesor LLM Benchmark21.3%
207Qwen3 4B Instruct 2507· Dubesor LLM Benchmark21.3%
208Qwen3.5 4B· Dubesor LLM Benchmark21.3%
209Hunyuan A13B Instruct· Dubesor LLM Benchmark21.1%
210GLM 4· Dubesor LLM Benchmark20.8%
211DeepSeek Coder V2 Instruct· Dubesor LLM Benchmark20.7%
212Gemma 4 E4B-itthinking· Dubesor LLM Benchmark20.7%
213Inflection 3 Pi· Dubesor LLM Benchmark20.6%
214Qwen3 Coder Flash· Dubesor LLM Benchmark20.6%
215Command A 03 2025· Dubesor LLM Benchmark19.7%
216GLM 4.7 Flash· Dubesor LLM Benchmark19.5%
217Mistral Nemo 12B Instruct· Dubesor LLM Benchmark19.5%
218QwQ 32B Preview· Dubesor LLM Benchmark19.1%
219Qwen2.5 Coder 7B Instruct· Dubesor LLM Benchmark18.7%
220Ministral 3 14B Instruct· Dubesor LLM Benchmark17.4%
221Gemma 4 E4B-it· Dubesor LLM Benchmark17.2%
222Mistral Small 3.2 24B Instruct· Dubesor LLM Benchmark17.2%
223Qwen 2.5 Coder 32B Instruct· Dubesor LLM Benchmark17.2%
224Claude Haiku 3.5· Dubesor LLM Benchmark17%
225Jamba 1.5 Large· Dubesor LLM Benchmark17%
226Command R 08 2024· Dubesor LLM Benchmark16.7%
227jamba-1.6-large· Dubesor LLM Benchmark15.1%
228Gemma 3n E4B IT· Dubesor LLM Benchmark14%
229GPT-4.1 nano· Dubesor LLM Benchmark13.9%
230Llama 2 70B Chat· Dubesor LLM Benchmark13.9%
231Jamba 1.5 Mini· Dubesor LLM Benchmark13.2%
232Nemotron 3 Nano 30B A3B· Dubesor LLM Benchmark13.2%
233Command R+· Dubesor LLM Benchmark12%
234Claude 3 Sonnet· Dubesor LLM Benchmark11.3%
235Olmo 2 32B Instruct· Dubesor LLM Benchmark11.2%
236Phi-3 Mini 4K Instruct· Dubesor LLM Benchmark11.2%
237Gemma 4 E2B-it· Dubesor LLM Benchmark10.9%
238GPT 5.4 Nano· Dubesor LLM Benchmark10%
239Gemma 3 4B IT· Dubesor LLM Benchmark9.5%
240LFM2-8B-A1B· Dubesor LLM Benchmark9.4%
241Qwen2.5 7B Instruct· Dubesor LLM Benchmark9%
242Mixtral 8x7B Instruct v0.1· Dubesor LLM Benchmark8.7%
243Gemma 2 2B IT· Dubesor LLM Benchmark8.1%
244Qwen3 1.7Bthinking· Dubesor LLM Benchmark8%
245Olmo 3 7B Instruct· Dubesor LLM Benchmark7.9%
246Phi-3 Medium 128K Instruct· Dubesor LLM Benchmark7.8%
247Claude 2.1· Dubesor LLM Benchmark7.3%
248Llama 3 8B Instruct· Dubesor LLM Benchmark7%
249Ministral 3 8b Instruct· Dubesor LLM Benchmark6.4%
250Gemma 4 E2B-itthinking· Dubesor LLM Benchmark4.9%

Showing 201–250 of 261 results.