Developed by Elon Musk’s artificial intelligence initiative xAI, the Grok 4 topped the analytical intelligence category, surpassing leading models in the industry. In the analysis index, which evaluates AI models using numerous technical and task-based metrics, the Grok 4 achieved the highest score with 73 points.

Grok 4 Tops Analytical Intelligence

The latest published table ranks prominent AI models based on various capabilities. In the AI index, which specifically evaluates analytics-focused capabilities, the Grok 4 achieved a score of 73. OpenAI’s o3-pro model ranked second with 71 points, while Google’s Gemini 2.5 Pro came in third with 70 points.

Grok 4’s leadership is attributed to its high accuracy, particularly in areas such as text reading, data interpretation, and complex scenario analysis. This development stands out as one of the areas where xAI has a technical advantage over its competitors in the AI race.

However, the same table doesn’t show Grok 4 topping the list in every metric. While the Grok 3 Mini Reasoning model tops the list in terms of cost per token, Grok 4 falls further behind in this category.

Similarly, Grok 4 appears to remain at average levels in performance-focused criteria such as throughput and latency. However, the Grok 4 model still ranks high in terms of total cost per task.

Elon Musk, with the data he shared, describes Grok 4’s progress as a “significant milestone.” It’s been commented that the emphasis xAI placed on analytics throughout its development process is being reflected in these results.

Analysis-based functions, such as task analysis, content filtering, and multi-layered inference, are beginning to play a strategic role in AI models. Technical updates and new benchmarks for Grok 4 are expected to continue to be released in the coming weeks.