xAI unveiled its Grok 3 models during a live broadcast with Elon Musk, who said they were “an order of magnitude more efficient than Grok 2.” The Grok 3 mini model can answer questions quickly, but it is not as accurate as other models in the family. Meanwhile, Grok 3 Reasoning and Grok 3 mini Reasoning models are able to simulate human-like reasoning when it comes to analyzing information required by the user.
Other examples of AI models capable of performing reasoning tasks are R1 by DeepSeek and o3-mini by OpenAI. According to TechCrunch, during the event, xAI claimed that Grok 3 Reasoning performs better than the best version of o3-mini on several benchmarks. Grok 3’s features will initially be available to subscribers paying for X’s Premium+ plan, which currently costs $40 per month in the U.S. (In December, X raised the price of its Premium+ tier from $16 to $22 – now, less than two months later, it’s almost double that.) They will also be available through an upcoming separate subscription option for the separate Grok app and Grok online. According to the leaked information, the subscription option will be called SuperGrok and will cost $30 per month.
With the Grok 3 models enabled, users will be able to ask the chatbot to “Think” if they want to use its mental abilities to solve math, science, and programming questions. For even more complex queries, they can use the “Big Brain” feature, which requires additional computation. The models’ ability to reason is at the heart of Grok’s new feature called DeepSearch, which xAI describes as “the next generation search engine.” DeepSearch will crawl the web and the social network X (formerly Twitter) to create a short summary for research queries.
In addition to launching the Grok 3 models, xAI also revealed during the event that the Grok app will receive a “voice mode” within a week, allowing it to synthesize voices to communicate with users. The Grok 2, the company’s older models, will be available as open source in the coming months.
