context.vn
Menu

arc Agi2

23 models evaluated

#ModelProviderTypeScore
1GPT-6 AstraOpenAIProprietary95%
2GPT-5.6 SolOpenAIProprietary92.5%
3Claude Opus 5AnthropicProprietary90.4%
4Claude Fable 5.1Anthropic · Closed90%
5GPT-5.5OpenAI · Closed85%
6GPT-5.6 TerraOpenAI · Closed83.9%
7GPT-5.4 ProOpenAI · Closed83.3%
8dots3-note PreviewDots Studio · Open weight81.4%
9Gemini 3.1 ProGoogle · Closed77.1%
10Claude Opus 4.7 (Adaptive)Anthropic · Closed75.8%
11GPT-5.4OpenAI · Closed74.0%
12Gemini 3.5 FlashGoogle · Closed72.1%
13Claude Opus 4.8Anthropic · Closed72.1%
14GPT-5.6 LunaOpenAI · Closed59.5%
15GPT-5.2 ProOpenAI · Closed54.2%
16Grok 4.20xAI · Closed53.3%
17GPT-5.2OpenAI · Closed52.9%
18Grok 4.5xAI · Closed52.6%
19Gemini 3 Pro Deep ThinkGoogle · Closed45.1%
20Muse SparkMeta · Closed42.5%
21Inkling-SmallThinking Machines Lab · Open weight40.1%
22Gemini 3 ProGoogle · Closed31.1%
23Claude Sonnet 4.5Anthropic · Closed13.6%