Skip to content

Benchmark

ARC-AGI v2

Scores are percentages.

Models measured
20
Scores recorded
20
Sources
1
IQ weight
1.37

Score distribution

45.4%
Mean
49.3%
Median
27.5%
Std dev
6.5%
10th percentile
77.1%
90th percentile
1.5%
Min
1.5%95%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.37 for chat models, measured over 20 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1GPT-6 Astra· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim95%
2GPT 5.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim85%
3Gemini 3.1 Pro Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim77.1%
4GPT 5.4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim73.3%
5Gemini 3.5 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim72.1%
6Claude Opus 4.6· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim68.8%
7Claude Sonnet 4.6· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim58.3%
8GPT-5.2 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim54.2%
9GPT 5.2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim52.9%
10Inkling 2 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim49.3%
11Muse Spark· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim42.5%
12Inkling Small· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim40.1%
13Claude Opus 4.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim37.6%
14Gemini 3 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim33.6%
15Gemini 3 Pro Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim31.1%
16Grok 4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim15.9%
17Claude Opus 4· LLM Stats data8.6%
18OpenAI o3· LLM Stats data6.5%
19Gemini 2.5 Pro· LLM Stats data4.9%
20Seed 2.0 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim1.5%

Data credits: Data by LLM Stats