According to OpenAI (@OpenAI), GPT-5.2 Pro is their smartest and most trustworthy AI model to date, offering significantly improved performance in complex domains such as programming and scientific research. This new iteration is positioned as the best model for assisting and accelerating scientists, highlighting its potential to streamline research workflows, automate code generation, and provide reliable support for challenging technical queries. The advancements in GPT-5.2 Pro are expected to drive AI adoption in R&D-intensive industries and elevate the use of large language models in enterprise and academic settings (source: OpenAI, 2025).

According to Demis Hassabis on Twitter, Gemini 2.5 Deep Think is now available for Ultra subscribers, offering a significant advancement in AI-driven creativity and planning. This model excels at solving complex problems by simultaneously considering, revising, and integrating multiple ideas, providing users with optimal solutions. Gemini 2.5 Deep Think is a faster version of the model that recently achieved gold-level performance at the International Mathematical Olympiad (IMO), highlighting its superior capabilities in handling intricate problem-solving tasks. Businesses and developers can leverage this technology for enhanced innovation, productivity, and competitive advantage in AI-powered applications (Source: @demishassabis, August 1, 2025).