Elon Musk's xAI launches Grok 4, promises ‘scientist-grade’ reasoning

Elon Musk’s artificial intelligence company xAI on Wednesday unveiled its latest AI model, Grok 4, touting faster reasoning, multimodal features and a redesigned interface, even as the company faces intensifying criticism over racist and antisemitic outputs from Grok’s earlier versions.

Musk and members of the xAI team introduced Grok 4 during a livestream on X, formerly known as Twitter, describing the release as an era of “Big Bang Intelligence.”

“We've run out of test questions to ask,” Musk said at the launch. “Reality is the ultimate reasoning test.”

Grok 4, trained on xAI’s Colossus supercomputer, is Musk’s most ambitious AI release to date. The model promises advanced scientist-grade reasoning, real-time web access through its DeepSearch tool, and upgraded multimodal capabilities that include images and potentially video.

A coding-focused variant, Grok 4 Code, aims to write, debug and explain code with greater accuracy, similar to GitHub Copilot and OpenAI’s Code Interpreter.

The launch comes at a time when xAI grapples with public outcry over Grok’s racist and antisemitic outputs.

On Tuesday, social media posts by Grok’s official account referred to Adolf Hitler as “history’s mustache man” and suggested he would be best placed to combat anti-white hatred, comments that drew widespread condemnation from users.

“We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts,” xAI said in a statement on X. “Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X.”

Musk, who has positioned xAI as an alternative to what he calls “woke” AI models such as ChatGPT and Google Gemini, has been largely silent on the controversy.

Since acquiring Twitter in 2022, Musk has reduced content moderation on the platform, a move that critics say has enabled the spread of extremist content and driven away advertisers.

The Grok 4 release intensifies competition with rivals OpenAI and Anthropic, whose flagship models GPT-5 and Claude 4 Opus have dominated headlines for their advanced reasoning, vision and real-time capabilities.