Elon Musk’s artificial intelligence firm xAI has unveiled the most recent iteration of its AI assistant, Grok. This release comes shortly after Grok’s antisemitic episode on X, where it propagated hate speech and lauded Hitler. In spite of this recent controversy, xAI asserts that the new Grok, version 4, is “the world’s most powerful AI model.”

In a livestream, xAI’s CEO Elon Musk hailed Grok 4 for its intellect, claiming it outshines “almost all graduate students, across all fields, at once,” although he conceded that it might “sometimes lack common sense.”

Musk also stated that Grok could possibly “uncover new physics” in the upcoming year, while confessing that its progress is “frankly, in certain aspects, somewhat frightening.”

Grok 4 is available for $30 monthly, with a more sophisticated version, Grok 4 Heavy, priced at $300 per month. This pricing approach reflects similar strategies by rivals such as OpenAI, Google, Anthropic, and Perplexity.

As per xAI, Grok 4 and Grok 4 Heavy exceed the performance of other models, including OpenAI’s o3, Google’s Gemini 2.5 Pro, and Claude 4 Opus in numerous AI benchmarks, although some critics argue that the comparisons might be skewed in favor of Grok.

The launch of Grok 4 transpired in a difficult context. Grok’s recent contentious incident on X was succeeded by the exit of X CEO Linda Yaccarino, while Musk has been involved in online altercations with President Donald Trump and is establishing a new political party in the U.S.

Disclosure: Ziff Davis, the parent company of Mashable, lodged a lawsuit against OpenAI in April, claiming copyright infringement in the training and operation of its AI systems.