OpenAI Unveils GPT-5.5: A Step Towards the AI Superapp

OpenAI announced on Thursday the launch of its latest artificial intelligence model, GPT-5.5. The company describes this model as the most intelligent and intuitive to use that it has ever created. Greg Brockman, co-founder and president of OpenAI, stated that this model represents a significant step towards the creation of a "superapp" for AI, a long-standing ambition for the company.

During a conference call with journalists, Brockman explained that GPT-5.5 constitutes a major advancement towards more agentic and intuitive computing. He said, "This model is a real step forward towards the type of computing we expect for the future — but it’s a step, and we expect to see many more in the future." According to him, GPT-5.5 is a faster and sharper thinker, requiring fewer tokens compared to its predecessor, model 5.4. This means that more cutting-edge AI is available for businesses and consumers, aligning with OpenAI's goals.

Brockman also discussed the vision of a "superapp," a versatile program akin to a Swiss army knife, that he and Sam Altman, co-founder of OpenAI, plan to launch. This superapp could combine ChatGPT, Codex, and an AI browser into a unified service aimed at assisting enterprise clients. This concept is also being explored by Elon Musk, who wants to transform X (formerly Twitter) into his own superapp.

OpenAI's development pace is rapid, with the launch of this new model occurring just a month after the previous one. The company has already released a model in December and another in November. Jakub Pachocki, chief scientist at OpenAI, indicated that the company expects to see significant improvements in the short term and extremely significant enhancements in the medium term. He added that the last two years have been surprisingly slow, but the pace should accelerate.

According to OpenAI, GPT-5.5 is designed to be useful across a wide range of categories. It performs particularly well in fundamental areas for businesses such as agentic coding and knowledge work, but also in more experimental AI applications like mathematics and scientific research. OpenAI has released data showing that GPT-5.5 consistently achieves higher scores than its previous models and those of competitors like Google and Anthropic, including Gemini 3.1 Pro and Claude Opus 4.5.

The rivalry between OpenAI and Anthropic is often at the center of discussions. A journalist asked during the press conference whether GPT-5.5 would have capabilities similar to Mythos, the cybersecurity tool recently announced by Anthropic. Mia Glaese, a member of OpenAI's technical team, responded that GPT-5.5 would have a significant impact on the company's approach to deploying its models for digital defense. She stated, "We have a strong and long-standing strategy for our approach to cybersecurity, and we have refined a sustainable approach to deploying models safely."

Mark Chen, research director at OpenAI, emphasized that GPT-5.5 is better at navigating computing tasks than its predecessors. He added that the model shows significant gains in scientific and technical research workflows, and it could really help expert scientists make progress. Chen also mentioned that GPT-5.5 could assist in drug discovery, an area that has garnered increased interest from the industry in recent years.

The GPT-5.5 model is widely available starting Thursday, according to OpenAI. The company states that the model is deployed for Plus, Pro, Business, and Enterprise users in ChatGPT, while 5.5 Pro is intended for Pro, Business, and Enterprise users.