A research preview of Operator, an agent that can use its own browser to perform tasks for you.

Operator is one of our first agents, which are AIs capable of doing work for you independently—you give it a task and it will execute it.

Operator is based on a new model we’re calling “computer-using agent” (CUA). CUA combines GPT-4o's vision capabilities with advanced reasoning through reinforcement learning. It’s trained to control a computer in the same way a human would—it looks at the screen, and uses a mouse and keyboard. The model still has limitations and will continue to evolve based on feedback. We plan to bring CUA to the API for developers soon.

For the first time, our models can take actions on the internet, so we did a lot of internal testing and external red teaming to help ensure Operator is safe to use.

Using custom instructions in Operator

Using saved prompts in Operator

Since this is a research preview, we are starting small—Operator will first be available to Pro users in the US at Eventually, this will be part of ChatGPT and available more broadly.

Want to help us continue building computer-using agents?

#AI agents: from tools to teammates. Software that can act, react, and interact with both humans and other agents will transform how businesses operate—and how they should invest in technology. This @EY_US article explains more. #ad

Wondering how restricted is it in real use cases with these constraints 👀

When o3?

"hi, I'm Sam and I lie to people for profit"