Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Exactly seven weeks after its launch GPT Chat 5.4,OpenAI is dropping A new family of models. ChatGPT 5.5 is now available to paid ChatGPT and Codex users, along with Think and Pro editions, with plans to bring it to the API soon. These models are designed for work—specifically, programming, computer use, and research.
OpenAI President Greg Brockman said GPT-5.5 is more intuitive than previous models and can do more with less human guidance.
“He can look at an unclear problem and decide what should happen next,” Brockman told reporters at a news conference. “It really feels like it’s laying the foundation for how we do computer work in the future, or how agent computing will work at scale.”
This is a generic model, and anyone can use it. But it’s likely to be most useful for people doing research or other intensive tasks, such as programming. He has agentic abilities, meaning he can complete tasks independently. It scored above GPT-5.4 in benchmarks that measure the model’s ability to use applications across your computer and solve mathematical problems.
OpenAI and other tech companies are trying to build models that act as true digital assistants, managing notifications and overseeing projects across your entire computer, not just one program. GPT 5.5 is the company’s biggest swing yet.
The company says GPT-5.5 has the strongest cybersecurity guarantees to date, and when it launches, it will be more conservative in the cybersecurity requests it allows. This is due to growing concerns that increasingly cyber-capable AI models could find vulnerabilities in our current internet architecture. That’s why they’re both anthropic Claude Mythos Preview And OpenAI 5.4-Cyber It was released to developers to root out issues before a potential public release.
Artificial intelligence in research
For OpenAI, research is a two-way street. GPT 5.5 is part of the foundation of the super application developed by the company It aims to build out of the Codex Alimentarius. Research on AI is also increasingly being conducted with the help of AI tools; Brockman said the company used 5.5 (and Codex) to help build itself during development.
AI has changed the way research is conducted over the past few years, and this field is not immune to the concerns of other industries Artificial intelligence may automatically put humans out of jobs in the future. While GPT-5.5 is OpenAI’s most capable research model to date, the company’s senior scientists and researchers aren’t worried about it being replaced by artificial intelligence.
Mark Chen, chief research officer at OpenAI, discussed pursuing a near-term goal of having humans serve as “coordinators” for research that gets significant help from AI.
Mia Glaze, vice president of research, said she hopes more capable AI models will help drive more research and raise “the threshold for what’s worth building.”
“I feel more productive because the challenge moves from figuring out implementation details, low-level abstractions, to higher-level goals,” said Jacob Paczucki, chief scientist at OpenAI. “It allows you to make progress much more quickly and spend your focus and energy figuring out ‘what things are important?’… There’s no benefit to automating things per se.”