The company OpenAI has unveiled GPT-5.5, its new artificial intelligence model optimized for handling complex tasks with a minimal set of instructions. According to the developers, the model supports scientific research, simplifies software development, and autonomously manages emails, spreadsheets, calendars, and other applications, executing user commands without constant supervision.
During a briefing reported by Bloomberg, OpenAI’s CEO Sam Brockman highlighted GPT-5.5’s ability to comprehend task objectives and effectively manage uncertainty, thereby improving intuitive interaction. The model is also noted for its superior performance in coding, one of the most lucrative areas in artificial intelligence. Furthermore, GPT-5.5 is set to form the foundation for a future integrated application combining chatbot features, programming tools, and web browsing capabilities.
OpenAI has also launched an early version of this model to expedite pharmaceutical drug development. GPT-5.5 is positioned as a partner for scientific research, capable of analyzing hypotheses and arguments, and, over time, partially automating scientific activities.
Amid growing market competition, OpenAI and Anthropic are actively vying to expand their client base, aiming to offset the significant investments they have made in advanced AI systems. Both companies focus on solutions for programming, cybersecurity, and scientific research, while also preparing for potential initial public offerings (IPOs) in 2026.
Anthropic has reinforced its position through high demand for programming tools, particularly its Mythos models, designed to identify vulnerabilities in critical software and currently available on a limited-access basis.
The GPT-5.5 model became accessible to ChatGPT Premium subscribers and professional developers using the Codex tool starting April 23, 2026.