Everyone’s switching to Claude but the smartest AI might surprise you (www.techradar.com)

Recent testing from OmniCalculator has upended the common perception of chatbots, revealing that xAI's Grok 4.2 excels in logic and problem-solving, outperforming both Claude and ChatGPT in mathematical reasoning. With a lower instability rate of 33.1% in complex scenarios, Grok demonstrates a stronger ability to maintain consistent conclusions compared to previous models like Claude and ChatGPT, which revise their answers roughly 60% of the time under similar conditions. This suggests that Grok has a distinct advantage in tasks requiring rigorous logical processes. Meanwhile, Anthropic's Claude remains a favorite for its superior writing quality and coherent response generation, making it particularly effective for longer, complex documents. Despite this, ChatGPT continues to be the most widely used AI chatbot, although its reputation as the "smartest" is challenged by the findings. The report highlights that while Grok is the best for technical problem-solving, the best chatbot for natural conversation and stylistic finesse may differ. As competition escalates in the AI space, specialization is likely to drive the development of these models, fostering a landscape where different AIs excel in various contexts rather than one dominating across the board.