- OpenAI brings Agent mode to its Mac app
- Agent mode is accessible from the toolbar menu
- Now you can leave long tasks running in the background while you work on something else
OpenAI has added its new Agent mode to its ChatGPT macOS app, and it’s available right now, if you’re a Plus subscriber. When you use ChatGPT through the Mac app, you’ll see Agent as one of the options in the toolbar that lives under the prompt window. Select it, and now you’re in Agent mode.
OpenAI launched Agent mode, or ChatGPT Agent, as it is also called, last week, and it’s a way to combine all the power of Deep Research with the agentic properties of its previous agent, called Operator.
During the launch event, Sam Altman and co showed off several uses for Agent, such as planning big events, like a wedding, or producing a presentation based on a wide range of data it has to go off and find.
OpenAI describes ChatGPT Agent as "ChatGPT that can think and act, proactively choosing from a toolbox of agentic skills to complete tasks for you using its own computer."
Alternative access
OpenAI also makes a version of its app for Windows users, but this hasn’t had Agent mode added yet.
We have noticed that some people don’t see the new Agent icon in their toolbar on the Mac app yet.
Presumably, this will be fixed as the change rolls out, but if you are using the Mac app and you don’t see an Agent button at the bottom of your interface, then don’t worry, you can still access Agent mode.
Simply type a / and a list of options will appear, one of which is 'Agent mode'.
The desktop app version of ChatGPT is more integrated into the operating system than simply using ChatGPT in a web browser.
For example, you can launch ChatGPT from any screen on your desktop with a keyboard shortcut. Use Option + Space on macOS or Alt + Space on Windows.
The Desktop version of the ChatGPT app also features Advanced Voice Mode, so you can chat with ChatGPT in real time using the microphone.
While we welcome the new ChatGPT Agent mode, it does open up your Mac to more vulnerabilities. This was something that the OpenAI team addressed at the launch, and Sam Altman said that there could be a new wave of threats that target AI agents who are simply trying to be helpful.
You might also like
- This Ozzy Osbourne tribute video may have 5.8M views, but it just shows everything wrong with AI slop (and what the hell is that music?)
- Sam Altman says there’s ‘Something about collectively deciding we're going to live our lives the way AI tells us feels bad and dangerous’ as OpenAI CEO worries about an AI-dominated future
- Sora 2 is coming, but it will have to dazzle viewers to beat Google's Veo 3 model
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.