Elon Musk’s artificial intelligence company, xAI, unveiled its new flagship model, Grok 4, yesterday (Wednesday) during a live-streamed event that showcased impressive technological capabilities alongside an unprecedented price tag. However, the grandiose announcement comes against the backdrop of a particularly tumultuous week for the Musk empire, raising poignant questions as to whether a leap in performance can compensate for the crises of trust and reputation that threaten to overshadow the ambitious venture’s future.

A Grandiose Announcement Against a Tumultuous Backdrop

During the launch event, held in a dramatic atmosphere, Musk sat alongside xAI executives and presented an uncompromising vision. “With respect to academic questions, Grok 4 is better than PhD level in every subject, no exceptions,” he declared. He added that, “At times, it may lack common sense, and it has not yet invented new technologies or discovered new physics, but that is just a matter of time.” These statements were intended to position Grok 4 as a direct, and even superior, competitor to the most advanced models on the market, such as OpenAI’s GPT-5, expected later this summer, and Google’s Gemini.

However, it is difficult to ignore the problematic timing of the announcement. Just hours earlier, Linda Yaccarino resigned from her role as CEO of X (formerly Twitter) after two years, a move that left the social network without clear leadership. Her resignation came only days after a serious incident in which Grok’s own automated X account posted antisemitic comments, which included criticism of “Jewish executives” in Hollywood and praise for Adolf Hitler. xAI was forced to limit the account’s activity, delete the offensive posts, and appeared to have removed a section from the bot’s system prompt that encouraged it not to shy away from making “politically incorrect” claims. During the launch event, Musk and his executives chose to almost completely ignore the incident, focusing instead on the new model’s performance data.

Impressive Benchmark Performance, But a More Complex Reality

According to data published by xAI, Grok 4’s performance on benchmarks indeed sets a new standard in the industry. The company launched two models: Grok 4, and a more advanced version named Grok 4 Heavy, which, according to Musk, works by splitting tasks among several “agents” that operate in parallel and compare their results “like a study group” to arrive at the best answer.

On the “Humanity’s Last Exam,” which measures the ability to answer thousands of questions across a variety of subjects, Grok 4 achieved a score of 25.4% without the use of external tools, surpassing Google’s Gemini 2.5 Pro (21.6%) and OpenAI’s o3 (21%). The Heavy version, with the assistance of tools, reached an impressive score of 44.4%, a significant lead over Gemini 2.5 Pro, which scored 26.9% in the same configuration. Furthermore, on the ARC-AGI-2 test, which consists of visual puzzles requiring pattern recognition, Grok set a new record with a score of 16.2% — nearly double that of the next leading commercial model, Claude Opus 4. These figures, on paper, place xAI at the technological forefront of the AI world.

A $300 Monthly Price Tag: Is the Enterprise Market Ready?

Alongside the new models, xAI launched a new and exceptionally expensive subscription plan: “SuperGrok Heavy” at a cost of $300 per month. These subscribers will get early access to the Heavy version and to new features that will be developed in the future. This price tag positions xAI as the most expensive artificial intelligence provider among the major companies, such as OpenAI, Google, and Anthropic, which offer similar premium tiers but at a lower cost.

The company hopes the early access will attract strategic customers, especially in light of the ambitious roadmap it presented: a coding model in August, a multi-modal agent in September, and a video generation model in October. xAI is releasing Grok 4 via an API to encourage developers to build applications on it, and notes that although its enterprise division is only two months old, it intends to collaborate with hyperscalers to expand the model’s availability. However, the combination of a high price and a controversial reputation may pose a significant hurdle in convincing enterprise clients to adopt the technology.

Conclusion and a Look Ahead: Between Technological Promise and the Test of Trust

The launch of Grok 4 illustrates the central dissonance that characterizes xAI and Elon Musk. On the one hand, this is a technology that, on paper, displays groundbreaking performance that threatens the hegemony of the established tech giants. On the other hand, the general conduct, from serious content incidents to ignoring them, combined with an aggressive pricing strategy, raises doubts about the company’s ability to become a major and reliable player in the enterprise market.

xAI’s greatest challenge is not technological, but rather one of marketing and ethics. The question is not whether Grok 4 is smarter than its competitors, but whether businesses will be willing to entrust their data and processes to a technology that has proven it can generate offensive content, and whose founder is constantly dealing with public controversies. The coming months will be critical in determining whether Grok 4’s impressive capabilities will succeed in convincing the market, or if its failures and controversies will leave it as a niche, glamorous product that businesses hesitate to adopt, with all its associated flaws.