Elon Musks Grok 4 Dethrones ChatGPT as Top AI
In a significant shift within the artificial intelligence landscape, Elon Musk's xAI has unveiled Grok 4, a model that has swiftly climbed the ranks to become the leading publicly available AI. Just over two years since its inception, Grok has surpassed established giants like OpenAI's ChatGPT, Google's Gemini, and models from Anthropic and Meta, marking a new chapter in the AI race. Adding to the excitement, Elon Musk announced that Grok will be integrated into Tesla vehicles starting next week.
Grok 4 Dominates the Benchmarks
The ascent of Grok 4 is not just hype; it's backed by data from independent, third-party testing platforms. The dramatic improvement from its predecessor is attributed to xAI's rapid expansion of its AI compute clusters, which have doubled to 200,000 GPUs with plans to reach one million.
To validate its capabilities, the xAI team engaged the creators of the rigorous ARC-AGI performance test. The results were compelling:
First, the facts: Grok 4 is now the top-performing publicly available model on ARC-AGI. This even outperforms purpose-built solutions submitted on Kaggle. Second, ARC-AGI-2 is hard for current AI models. To score well, models have to learn a mini-skill from a series of training examples, then demonstrate that skill at test time. The previous top score was ~8% (by Opus 4). Below 10% is noisy. Getting 15.9% breaks through that noise barrier, Grok 4 is showing non-zero levels of fluid intelligence.
Another tester, Artificial Analysis, corroborated these findings, stating, "We have run our full suite of benchmarks and Grok 4 achieves an Artificial Analysis Intelligence Index of 73, ahead of OpenAI o3 at 70, Google Gemini 2.5 Pro at 70, Anthropic Claude 4 Opus at 64 and DeepSeek R1 0528 at 68."
Musk's Vision and Current Limitations
During the Grok 4 release presentation, Elon Musk made characteristically bold claims, suggesting the model is now smarter than all graduate students combined. He projected that by next year, Grok 4 could independently discover new technologies, such as novel medicines or engineering breakthroughs.
However, Musk also tempered expectations, admitting that Grok will remain weak in image recognition for another month or so. He also addressed a recent controversy over biased answers, explaining that, "when Grok goes far wrong, that is usually due to something foolish we did, like a bad system prompt, or placing too much weight on biased sources."
The Price of Top-Tier AI
The launch of Grok 4 also introduces a new premium pricing structure designed to monetize its advanced capabilities. For the most demanding users, xAI is offering a "SuperGrok Heavy" tier at a substantial $300 per month. This plan includes everything from the lower tier, plus access to the Grok 4 Heavy platform which provides higher rate limits and early access to new features.
A more accessible "SuperGrok" tier, which grants initial access to Grok 4, will be included for all X Premium+ subscribers at $30 per month. Meanwhile, the previous version, Grok 3, will remain free for public use.
