Last updated: September 29, 2026
The quick answer
At DevDay on September 29, 2026, OpenAI launched “dots,” always-on AI agents that run on their own cloud computer and keep working on tasks while you are away. The first dot is included with ChatGPT Pro and Business Premium in eligible markets, with a beta for Enterprise, Edu and Healthcare workspaces; Free and Plus users do not get one. It is powerful, but it is built for people who will actually hand off ongoing work and review what it does.
What happened
OpenAI used its annual developer conference to announce more than 20 launches, and dots were the headline for everyday ChatGPT users. Unlike a normal chat, which only works while you are typing, a dot is a persistent agent. It runs on GPT-6 Astra, OpenAI’s flagship model, has its own cloud computer and browser, and can connect to more than 4,000 apps through OpenAI’s plugin ecosystem.
You give a dot goals rather than one-off prompts. Coverage of the launch describes dots doing background research while you are inactive, juggling several projects, learning your preferences from feedback, and preparing things like an overdue invoice, then waiting for your approval before sending it. You can talk to your dot in ChatGPT on desktop, web and mobile, in Slack and Microsoft Teams, and by voice, with texting announced as coming soon.
OpenAI also introduced GPT-6.1 Sol, a lower-cost model it describes as coming close to Astra-level performance, aimed at coding, computer use and professional tasks.
What changes for you
| Plan | Dots access |
|---|---|
| ChatGPT Free and Plus | Not included |
| ChatGPT Pro | First dot included at no extra cost, rolling out in eligible markets |
| ChatGPT Business Premium | First dot included, rolling out in eligible markets |
| Enterprise, Edu and Healthcare | Beta, once a workspace admin turns it on |
Two details matter for heavy users. Chatting with your dot does not count toward your ChatGPT usage limits, but work it kicks off in Codex or ChatGPT Work does. And because a dot acts on connected accounts, its usefulness depends on how much access you give it.
How to set up a dot safely
- Start from a computer. Your first dot is created from ChatGPT on desktop. After that, you can message it from the ChatGPT mobile app.
- Connect only what it needs. Link the specific apps tied to the goal you are giving it, not every account you have. You can add more later.
- Write custom rules. Dots let you allow, block or require approval for specific actions. Require approval for anything that sends messages, spends money or shares files.
- Keep it off your own PC at first. By default a dot works on its separate cloud computer. Connecting it to your personal computer is optional, so hold off until you trust how it behaves.
- Watch it work. You can follow a dot’s progress and see what it is doing on its cloud computer. Check in often during the first week.
- Review before anything important goes out. OpenAI’s own guidance is blunt: “Dots can still make mistakes, so always review consequential work.”
OpenAI says certain sensitive actions, such as changing a password, will always require you to do them yourself. That guardrail helps, but it does not replace your own review, especially while always-on agents are this new.
Is it worth it?
For most people, not on its own. Dots are only available on OpenAI’s higher-priced plans, so upgrading from Plus just to get one makes sense only if you have recurring, multi-step work that eats real time each week, such as tracking invoices, monitoring a project channel or pulling research together. If you mainly use ChatGPT for questions and writing help, your current plan still covers what you need, and it makes sense to wait for early reviews before paying more.
Who it affects and who it doesn’t
- Affects: ChatGPT Pro and Business Premium subscribers in eligible markets, plus Enterprise, Edu and Healthcare users whose admins enable the beta.
- Indirectly affects: coworkers in Slack or Teams, who may start receiving messages or requests prepared by someone’s dot.
- Does not affect: Free and Plus users, who cannot create dots.
Google is heading in a similar direction with reusable AI instructions; see our guide to Gemini Gems becoming skills for what changes there.
FAQ
Can a dot act without my permission?
It can take some actions on its own within the rules you set, which is the point of an always-on agent. You decide which actions it can take freely, which need your approval and which are blocked, and sensitive tasks like password changes always stay with you.
Do conversations with a dot use up my ChatGPT limits?
No. Chatting with a dot does not count toward ChatGPT usage limits. Tasks it runs through Codex or ChatGPT Work do count.
Can I use a dot on my phone?
Yes, once it is set up. You create your first dot from a computer, then you can message it in the ChatGPT mobile app, as well as in Slack and Microsoft Teams.
Sources
- 9to5Google: OpenAI launches Dots, new always-on agents you can assign tasks to
- The Next Web: OpenAI launches dots, always-on AI agents with their own cloud computers
- Decrypt: OpenAI gave AI agents their own computers at DevDay 2026
- Digit: OpenAI dots announced at DevDay 2026
- BGR: Everything OpenAI announced at DevDay 2026
Leave a Reply