Skip to content

Benchmark

Lmca

Scores are percentages.

Models measured
141
Scores recorded
205
Sources
1
IQ weight
1.56

Score distribution

41.6%
Mean
44.5%
Median
18.1%
Std dev
15.2%
10th percentile
64.8%
90th percentile
2.8%
Min
2.8%70.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.56 for chat models, measured over 141 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101Claude Haiku 5.5low· Epoch AI Benchmarking Hub data44.8%
102GPT-6 Lunamax· Epoch AI Benchmarking Hub data44.5%
103Claude Opus 4.5high· Epoch AI Benchmarking Hub data44.5%
104GPT-5.6 Lunamedium· Epoch AI Benchmarking Hub data44.3%
105Qwen3.7 Max· Epoch AI Benchmarking Hub data44%
106GPT 5.2xhigh· Epoch AI Benchmarking Hub data43.9%
107GPT 5.1high· Epoch AI Benchmarking Hub data43.9%
108GPT-5.6 Lunalow· Epoch AI Benchmarking Hub data43.8%
109Gemini 3 Flashhigh· Epoch AI Benchmarking Hub data43.1%
110Qwen3.6 Max Preview· Epoch AI Benchmarking Hub data42.5%
111DeepSeek V4 Flash 0731max· Epoch AI Benchmarking Hub data41.7%
112GPT-5.6 Lunanone· Epoch AI Benchmarking Hub data41.4%
113Qwen 3.8 27B· Epoch AI Benchmarking Hub data41.4%
114DeepSeek V4 Promax· Epoch AI Benchmarking Hub data41.2%
115GPT-5.4 minixhigh· Epoch AI Benchmarking Hub data40.8%
116GPT 5high· Epoch AI Benchmarking Hub data40%
117OpenAI o3high· Epoch AI Benchmarking Hub data39.7%
118Claude Haiku 5.5none· Epoch AI Benchmarking Hub data39.3%
119Gemma 4 31B IT· Epoch AI Benchmarking Hub data39.3%
120Claude Sonnet 4.5· Epoch AI Benchmarking Hub data38.8%
121Grok 4.20· Epoch AI Benchmarking Hub data38.7%
122O3 Prohigh· Epoch AI Benchmarking Hub data38.5%
123Grok 4.3high· Epoch AI Benchmarking Hub data38.3%
124Qwen3.5 397B A17B· Epoch AI Benchmarking Hub data37.9%
125Gemini 3.5 Flash Litehigh· Epoch AI Benchmarking Hub data37.6%
126Inkling FP4xhigh· Epoch AI Benchmarking Hub data37.6%
127Qwen3.7 Plus· Epoch AI Benchmarking Hub data37.6%
128Claude Opus 4· Epoch AI Benchmarking Hub data37.4%
129Kimi K2.6· Epoch AI Benchmarking Hub data37.3%
130Claude Opus 4.1· Epoch AI Benchmarking Hub data37.1%
131Nemotron 3 Ultra· Epoch AI Benchmarking Hub data36.9%
132GPT 5.4 Nanoxhigh· Epoch AI Benchmarking Hub data36.9%
133Qwen3.5 Plus· Epoch AI Benchmarking Hub data36.4%
134DeepSeek V4 Flashmax· Epoch AI Benchmarking Hub data35.9%
135Mistral Large 4· Epoch AI Benchmarking Hub data35.8%
136Gemini 3.1 Flash Litehigh· Epoch AI Benchmarking Hub data35%
137Gemini 2.5 Pro· Epoch AI Benchmarking Hub data34.8%
138Qwen3.6 27B· Epoch AI Benchmarking Hub data34.5%
139GPT-5 minihigh· Epoch AI Benchmarking Hub data34.2%
140Qwen3.5 27B· Epoch AI Benchmarking Hub data34%
141MiniMax M3· Epoch AI Benchmarking Hub data33.7%
142Qwen3.6 Plus· Epoch AI Benchmarking Hub data33.1%
143Qwen3.5 122B-A10Bnone· Epoch AI Benchmarking Hub data32.2%
144Qwen3.6 Flash· Epoch AI Benchmarking Hub data31%
145Claude Haiku 4.5· Epoch AI Benchmarking Hub data30.9%
146Qwen3.6 35B A3B· Epoch AI Benchmarking Hub data29.7%
147Gemma 4 26B A4B· Epoch AI Benchmarking Hub data29.7%
148MiMo V2.5 Pro· Epoch AI Benchmarking Hub data29.5%
149Qwen3.5 35B A3B· Epoch AI Benchmarking Hub data29.5%
150Qwen 3 235B A22B Thinking 2507· Epoch AI Benchmarking Hub data29.3%

Showing 101–150 of 205 results.

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