The Number 1 AI King Isn’t ChatGPT-Google-Microsoft, It Turns Out This Is It

Jakarta, CNBC Indonesia –Grok 4 has become the world’s leading AI model. In a study, xAI’s benchmark model surpassed OpenAI and Gemini.

The report, based on a post by the independent analysis account Artificial Analysis, found that Grok 4 achieved an Artificial Analysis Intelligence Index score of 73, becoming the first to lead the AI frontier.

This score surpasses OpenAI o3 at 70, Google Gemini 2.5 Pro at 70, Anthropic Claude 4 Opus at 64, and DeepSeek R1 0528 at 68.

“Grok 4 leads not only in the Artificial Analysis Intelligence Index but also in the Coding Index and Math Index,” the report stated, quoted on Friday (July 11, 2025).

This achievement also surpasses Grok 3. Previously, the model had competitive scores with the latest models from OpenAI, Anthropic, and Google.

“However, Grok 4 marks the first time our Intelligence Index has shown xAI to be in first place,” the report stated.

Furthermore, Grok 4 recorded slightly higher output token usage than similar models, resulting in higher fees compared to the price per token.

Grok 4 testing used the xAI API. However, it’s important to note that Grok 4 for social media X is somewhat different from the model available through the API.

The consumer application version for a typical large language model has instructions and logic that change the model’s style and behavior.

Grok 4 is a reasoning model, so it thinks before answering. The xAI API, on the other hand, does not share the generated reasoning tokens.

SOURCE : CNBC INDONESIA