xAI, the artificial intelligence company founded by Elon Musk, has unveiled its latest chatbot model, Grok 3. Musk announced that Grok 3 boasts greater capabilities than its predecessor, Grok 2, continuing the path towards creating AI that seeks the truth of the universe, even if that truth may sometimes stray from political correctness.
Grok 3 is comprised of multiple sub-models tailored for specific tasks, such as Grok 3 mini, a compact version emphasizing quick responses with minimal processing resources, and Grok 3 Reasoning, a step-by-step reasoning model still in beta testing. In mathematical tests like AIME’24, Grok-3 outperformed other models like Gemini-2 Pro, DeepSeek-V3, and GPT-4o, with Grok-3 mini also scoring impressively. Similarly, in step-by-step reasoning tests like AIME 2025, Grok-3 Reasoning and Grok-3 mini Reasoning scored higher than comparable models like o3mini, o1, or DeepSeek-R1.
Premium+ customers will have early access to Grok 3, with xAI also introducing a new subscription tier called SuperGrok, offering enhanced features like DeepSearch for in-depth research inquiries and unlimited image generation at $30 per month or $300 annually.
Future plans include Musk’s announcement of voice-activated functionality for Grok 3 in the coming weeks, followed by an API for Enterprise customers and the integration of DeepSearch features within a few months. xAI will also open-source older Grok 2 models.
TLDR: xAI introduces Grok 3 with advanced features, outperforming previous models in tests, and plans for voice commands, API access, and open-sourcing older models in the near future.
Leave a Comment