Quick Answer
OpenAI Dots is a new type of AI helper that lives inside ChatGPT and never really switches off. Unlike a normal chat that stops the moment you close the app, a dot keeps working in the background — checking your apps, doing research, and finishing small tasks — then comes back to you with results. It launched in late September 2026 and runs on OpenAI’s newest model, GPT-6 Astra. Your first dot comes free with a ChatGPT Pro or Business Premium plan.
Have you ever wished you could have an assistant that would continue to work on your to-do list while you slept? OpenAI Dots is the innovation from OpenAI that does exactly that. It’s also one of the most significant developments for the company this year, fundamentally changing how people can use ChatGPT in the workplace. Let’s take a look at what OpenAI Dots actually is in simple terms.
What Is OpenAI Dots, Exactly?
A dot is an individualized artificial intelligence agent that resides in the user’s ChatGPT account. It is similar to a virtual employee. The user is required to provide a name and design for their dot, and assign it a task. After that, it will continue to perform the duties even when the user is not interacting with the bot.
Each dot has its own virtual computer in the cloud and a virtual browser. This way, the dot can access the internet, read articles, fill out forms, and perform other tasks. They can also be integrated into the apps that the user authorizes, including Slack, Microsoft Teams, Google Drive, and over 4000 other software.
How Is a Dot Different From Normal ChatGPT?
A normal ChatGPT conversation is a bit like a phone call — once you hang up, nothing else happens. A dot is more like texting a real assistant: you can walk away, and it keeps going.
Here’s the simplest way to picture the difference:
- Regular ChatGPT: You ask a question, it answers, the conversation ends.
- A Dot: You give it a goal, it keeps working toward that goal over hours or days, and it messages you when it needs help or has something to show you.
Key Features of OpenAI Dots
It Has Its Own Computer and Browser
Each dot runs on its own cloud computer, separate from your laptop or phone. It can browse websites and use apps on its own. If you want, you can also let it access files on your own computer — but that’s switched off unless you turn it on yourself.
It Remembers Things and Does Research Without Being Asked
A dot accumulates memory over time, similarly to how ChatGPT does it now, but takes it one step farther. OpenAI refers to this as “proactive research”: your dot can perform research (while you are waiting) on the apps it has accessed, uncovering valuable insights that it can flag for your attention, even if you do not ask it to do so directly. During this research phase, it can only observe: it cannot modify or reply to any messages.
You Set the Rules
This is the part worth paying close attention to. You decide how much freedom your dot has using something called Custom Rules. For any task, you can tell it to:
- Just do it, no questions asked
- Do it, but only if you already said it was okay
- Ask first before doing it
- Leave it for you to handle yourself
Some things, like changing a password, always need your approval — no setting can turn that off.
You Can Reach It From Almost Anywhere
You can message or talk to your dot through ChatGPT on your desktop, the web, or your phone. It also works inside Slack and Microsoft Teams, and text message support is being tested. Wherever you message it, it remembers the full conversation history.
What Can OpenAI Dots Actually Do? Real Examples
OpenAI shared a few examples that show how this plays out in daily life:
- Fixing small bugs: A dot can watch for repeated customer complaints, figure out the fix, build it, test it, and hand you a ready-to-review pull request.
- Marketing updates: When a product changes, it can quietly update the marketing copy and related documents to match.
- Research work: As new data comes in, it can rerun your analysis, double-check odd results, and flag anything worth your attention.
- Sales support: It can check a client’s requirements against your product and even put together a quick demo.
- The small stuff nobody gets around to: One early user said their dot noticed they’d forgotten to send an invoice, prepared it, and sent it the moment they said “go ahead.”
That last example says a lot. Dots aren’t just for huge automated workflows — they’re built for the small, annoying tasks that pile up because nobody has time for them.
OpenAI Dots vs. ChatGPT vs. Codex
| Feature | Regular ChatGPT | Codex | OpenAI Dots |
| Keeps working after you leave | No | Limited | Yes |
| Has its own browser/computer | No | Sometimes | Yes |
| Works across Slack & Teams | No | No | Yes |
| Learns and remembers over time | Some | Limited | Yes, built-in |
| Needs a new message to start | Yes | Yes | No |
Price and Availability
The good news: your first dot is free if you already pay for ChatGPT Pro or Business Premium. There’s no extra subscription. Business Premium users get access everywhere it’s offered, while Pro users in the UK, Switzerland and the European Economic Area aren’t able to access it yet. Enterprise and education accounts can try it out once an administrator has enabled it.
What OpenAI Dots Can’t Do Yet
No tool is perfect on day one, and Dots has a few clear gaps right now:
- You can’t create a dot on your phone — setup needs a desktop computer.
- It can’t call you first; you have to reach out to it.
- It doesn’t have its own email address yet.
- Deleting its memory means deleting the whole dot — there’s no “forget just this one thing” option.
- OpenAI hasn’t yet said what it will cost to use after the free first month.
Is OpenAI Dots Safe?
OpenAI notes that the model which was used to create Dots was trained to refuse any harmful requests, including those related to hacking or conducting dangerous experiments. The company undertook several safety tests to ensure that dots would not be able to fool users into requesting something which goes beyond the technology’s limits. It is noteworthy that all dangerous actions always require permission from the user, regardless of their settings. At the same time, OpenAI admits that dots can make mistakes, so it is always necessary to make sure that everything is in order before sending a request.
How to Set Up Your First Dot
- Open the ChatGPT desktop app or ChatGPT.com on a desktop browser.
- Follow the prompts to create and name your first dot.
- Connect the apps you want it to use, like Slack or Google Drive.
- Set your Custom Rules so it knows what it can do without asking.
- Give it a first task and see what comes back.
Frequently Asked Questions
What is OpenAI Dots in simple terms?
It’s an AI helper inside ChatGPT that keeps working on tasks even after you stop chatting with it, instead of needing a new message every time.
Is OpenAI Dots free?
Your first dot comes included if you already have a ChatGPT Pro or Business Premium plan. There’s no separate fee for it at launch.
Can OpenAI Dots access my personal computer?
Only if you turn that on yourself. By default, a dot works on its own separate cloud computer, not your device.
Is OpenAI Dots available in the UK?
Not yet for Pro users. At launch, Pro access does not include the UK, Switzerland, or the European Economic Area, though Business Premium customers in those regions can use it.
What’s the difference between OpenAI Dots and Codex?
Codex is built mainly for coding tasks you start and manage yourself. A dot is a broader assistant that keeps working on all kinds of tasks without needing to be restarted.
