Skip to content

Benchmark

ARC-AGI

Scores are percentages.

Models measured
102
Scores recorded
295
Sources
2
IQ weight
2.26

Score distribution

64.1%
Mean
75%
Median
31.2%
Std dev
14.5%
10th percentile
97%
90th percentile
0%
Min
0%98.5%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.26 for chat models, measured over 102 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
151GPT-5.6 Sollow· Epoch AI Benchmarking Hub data74.5%
152GPT-6 Lunaxhigh· Epoch AI Benchmarking Hub data73%
153GPT 5.1high· Epoch AI Benchmarking Hub data72.8%
154GPT 5.2medium· Epoch AI Benchmarking Hub data72.7%
155GPT-6 Sollow· Epoch AI Benchmarking Hub data72.2%
156Claude Opus 4.516k· Epoch AI Benchmarking Hub data72%
157GLM 5.3 Flashhigh· Epoch AI Benchmarking Hub data71.8%
158GPT-6 Lunahigh· Epoch AI Benchmarking Hub data70.3%
159GPT-5 Prohigh· Epoch AI Benchmarking Hub data70.2%
160GPT-5 Pro· Epoch AI Benchmarking Hub data70.2%
161Claude Sonnet 5.5medium· Epoch AI Benchmarking Hub data70%
162Qwen 3.8 27Blow· Epoch AI Benchmarking Hub data69.2%
163Qwen 3.8 27Bmedium· Epoch AI Benchmarking Hub data68.7%
164GPT 5.4low· Epoch AI Benchmarking Hub data68.2%
165Doubao Seed 1.8· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim67.9%
166Claude Sonnet 5.5low· Epoch AI Benchmarking Hub data67%
167Inkling Smallmedium· Epoch AI Benchmarking Hub data67%
168Muse Glimmer 30B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.9%
169Grok 4· Epoch AI Benchmarking Hub data66.7%
170GPT 5high· Epoch AI Benchmarking Hub data65.7%
171Kimi K3low· Epoch AI Benchmarking Hub data65.7%
172Kimi K2.5· Epoch AI Benchmarking Hub data65.3%
173Claude Sonnet 4.532k· Epoch AI Benchmarking Hub data63.7%
174GPT-5.4 minixhigh· Epoch AI Benchmarking Hub data63.7%
175MiniMax M2.5· Epoch AI Benchmarking Hub data63.7%
176Grok 4.7low· Epoch AI Benchmarking Hub data63.3%
177GPT-6 Lunamedium· Epoch AI Benchmarking Hub data61%
178OpenAI o3high· Epoch AI Benchmarking Hub data60.8%
179GPT-5.6 Terralow· Epoch AI Benchmarking Hub data60.2%
180O3 Prohigh· Epoch AI Benchmarking Hub data59.3%
181Claude Opus 4.58k· Epoch AI Benchmarking Hub data58.7%
182o4 Minihigh· Epoch AI Benchmarking Hub data58.7%
183Muse Spark Open 816B A42Blow· Epoch AI Benchmarking Hub data58.5%
184GPT-5.4 minihigh· Epoch AI Benchmarking Hub data58%
185GPT 5.1medium· Epoch AI Benchmarking Hub data57.7%
186Claude Haiku 5.5medium· Epoch AI Benchmarking Hub data57.3%
187DeepSeek V3.2· Epoch AI Benchmarking Hub data57%
188O3 Promedium· Epoch AI Benchmarking Hub data57%
189GPT-5.6 Lunamedium· Epoch AI Benchmarking Hub data56.5%
190GPT 5medium· Epoch AI Benchmarking Hub data56.2%
191GPT 5.2low· Epoch AI Benchmarking Hub data55.7%
192GPT-5 minihigh· Epoch AI Benchmarking Hub data54.3%
193OpenAI o3medium· Epoch AI Benchmarking Hub data53.8%
194Gemini 3.5 Flash Litehigh· Epoch AI Benchmarking Hub data53.5%
195Inkling Smalllow· Epoch AI Benchmarking Hub data52.5%
196GPT 5.4 Nanoxhigh· Epoch AI Benchmarking Hub data51.5%
197Claude Haiku 5.5low· Epoch AI Benchmarking Hub data50.3%
198LongCat-Flash-Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim50.3%
199Gemini 3.5 Flashminimal· Epoch AI Benchmarking Hub data48.8%
200Grok 4 Fast· Epoch AI Benchmarking Hub data48.5%

Showing 151–200 of 295 results.

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · Data by LLM Stats