OpenAI has expanded its family of artificial intelligence (AI) models with the introduction of the mini and nano versions of GPT-5.4, which bring advanced capabilities such as multimodal understanding and web searching to more efficient and faster models. The mini version, which is free to use, allows users to enjoy some of the advanced functions Found in the full GPT-5.4 model, including control of computer applications autonomously.

Sam Altman, CEO of OpenAI, initially presented GPT-5.4 in early March, claiming it as the first model to offer native computing capabilities. This advancement means that AI agents can autonomously execute workflows across applications, providing unprecedented control over computer tasks and effectiveness in performance. The introduction of the mini and nano versions is strategically aimed at increasing accessibility to these advanced tools and democratizing access to AI capabilities for broader audiences.

The mini version specifically boasts enhancements in programming and task management, making it suitable for handling large workflows efficiently. By incorporating some advanced functionalities from the flagship model while remaining lightweight, these new versions represent OpenAI's commitment to improving user experience and expanding the practical applications of AI in everyday tasks, positioning itself prominently in the competitive AI landscape.