Elon Musk's xAI company introduced its latest artificial intelligence model, Grok 3, on February 17, 2025, which the company claims surpasses leading AI chatbots, including ChatGPT and DeepSeek. The model was developed on the Memphis "Colossus" supercomputer, which utilizes 200,000 Nvidia GPUs, providing ten times more computing capacity compared to the previous version.
The Grok 3 family includes multiple models, among them the faster but less accurate Grok 3 mini, as well as the Grok 3 Reasoning and Grok 3 mini Reasoning variants, which are specialized for solving complex mathematical, scientific, and programming tasks. During development, special emphasis was placed on training with synthetic data, which, according to xAI, resulted in better reasoning capabilities. This is supported by Musk's statement at the Dubai World Government Summit: Grok 3 has the strongest reasoning capabilities and outperformed every model released so far that we know of during testing.
The new model will first be available to X platform Premium+ subscribers (for $50 per month). At the same time, additional features, such as DeepSearch and unlimited image generation, will be accessible through the SuperGrok subscription (priced at $30 per month or $300 per year). According to the company's plans, Grok 3 will be capable of voice-based interaction one week after release, while enterprise API access will become available within a few weeks. Additionally, xAI announced that they will make Grok 2's source code public once Grok 3 is entirely stable.
Sources:
1.
India Today Tech
2.
Brian Buntz
3.
Rebecca Rommen
4.
Isobel O’Sullivan
5.
Dylan Butts
