OpenAI has introduced Dots, a new kind of AI agent that keeps working in the background even when you are not chatting with it. CEO Sam Altman launched the product at OpenAI DevDay on Tuesday.
Dots are not a new chatbot window. They are meant to pursue goals you set continuously, with minimal oversight, and to run independently of any specific hardware or interface. The catch is the price of entry: they are limited to paid plans, and not every region gets them yet.
OpenAI announced dots in a press release on Tuesday, Sept. 29, and began rolling them out in ChatGPT the same day. You can start with a primary dot, give it a name, and make it your own. OpenAI says it expects teams of dots working together over time.
Key Details
- Dots run on GPT-6 Astra, have their own cloud computer, learn from feedback, and can work around the clock.
- They connect to more than 4,000 apps through OpenAI’s plugin ecosystem.
- You can message a dot through Slack, Teams, and other workplace platforms. Text message support is listed as coming soon.
- Dots also work with ChatGPT Voice Mode, so you can hand off tasks by talking, much like a call with a human assistant.
OpenAI gave two examples of use. A dot can handle small jobs like booking a table, or take on bigger work in the way a high-agency engineer or chief of staff would. In one demo scenario, a dot spotted a software bug alert in Slack and started digging into it.
Why It Matters
Much of this was already possible through tools like Codex. Dots bundle those abilities into one package built around independent action. That shifts AI from something you prompt to something you delegate to.
Safety and Control
OpenAI says you stay in charge. You set what a dot can do on its own, when it must ask first, and what it should never do. Dots are also trained not to handle certain sensitive tasks, such as changing passwords or moving money between accounts.
OpenAI is open about the risk. In its safety post, the company admits an agent’s mistakes can reach beyond a chat, such as editing the wrong file or sharing private information.
Price and Availability
Dots have no separate price. Your first dot is included in a Pro or Business Premium plan at no extra cost.
| Plan | Monthly price | Dots access |
| ChatGPT Pro 100 | $100 | First dot included |
| ChatGPT Pro 200 | $200 | First dot included |
| ChatGPT Pro 500 | $500 | First dot included |
| Business Premium | $125 per seat (monthly billing) | First dot included |
| Free, Go, Plus | Various | Not listed as eligible |
Free, Go, Plus, and Business Standard are not listed as eligible launch plans. Enterprise, Edu, and Healthcare workspaces have it in beta, switched off by default.
Read Also: OpenAI Launches GPT-6 Astra, Says We’re Now in the “AGI Era”
Region matters too. Pro subscribers in the European Economic Area, Switzerland, and the UK cannot use dots yet, while Business Premium customers can in all supported regions. OpenAI has not said whether India is covered. It only says “eligible markets,” so check your own account before paying.
Extra dots are planned. OpenAI has not published a price for them. Per-plan usage terms will follow after an initial launch window.
Industry Impact
The timing is not accidental. The launch came three weeks after Meta introduced its personal AI agent, Muse, and a day after Instinct, another personal-agent startup, raised $1 billion at a $10 billion valuation. The race to build the assistant that lives in your daily workflow is now wide open.
Final Thoughts
Dots is making a serious bet on agents that act for you, rather than simply talking to you. For most people, the $100 monthly entry price and regional limits put them out of reach for now. Watch what OpenAI announces on extra dots and usage terms, because that will decide whether this stays a power-user tool.
