On July 10, Artificial Analysis unveiled its benchmark results, revealing that Musk's Grok 4 model has attained an intelligence index of 73. This achievement places Grok 4 ahead of OpenAI's o3 model, which scored 70, as well as Google's Gemini 2.5 Pro model, which also received a score of 70. Furthermore, Grok 4 outperformed Deepseek's R1-0528 model with a score of 68 and Anthropic's Claude 4 Opus model, which scored 64.