OpenAI launches Dots: always-on agents powered by GPT-6 Astra

OpenAI announced Dots on September 29, 2026: a new class of always-on agents built on top of ChatGPT. Unlike a chat assistant that answers one question at a time, a Dot is designed to take ongoing responsibility for work — starting projects, following through, and reporting back.

The one-paragraph summary

A Dot is a persistent agent with its own cloud computer and browser, powered by GPT-6 Astra. It can write and test code, connect to more than 4,000 apps through OpenAI’s plugin ecosystem, learn from your feedback over time, and keep working toward your goals around the clock. You reach it through ChatGPT (desktop, web, mobile), and it can also message you.

What actually shipped

• Your primary Dot — you give it a name and make it yours. The first Dot is included with eligible Pro and Business Premium plans.
• Own cloud computer and browser — you can open the Dot’s computer at any time to inspect its work.
• Plugin ecosystem — connections to 4,000+ apps, managed through your existing ChatGPT app controls.
• Multi-channel access — ChatGPT on desktop, web, and mobile, plus Slack and Teams, with texting (iMessage/RCS) on a limited waitlist.
• Specialist Dots (preview) — organizational agents with their own identity, credentials, and access to systems of record. OpenAI is running focused enterprise pilots.
• Microsoft Agent 365 integration — work to let businesses govern specialist Dots through Microsoft’s existing tools.

Why “always-on” is the point

The defining feature is not any single capability but persistence. A Dot can hold several projects at once, take new tasks without you juggling separate threads, and act while you are away. OpenAI describes the intended experience as work that comes back “done the way you would do it, sometimes before you even think to ask.”

That framing matters for anyone evaluating it: the value shows up over time as the Dot learns your preferences and standards, not in a single prompt.

What we still don’t know

OpenAI has not published detailed throughput limits per Dot, the exact markets in the first rollout wave beyond “eligible markets,” or final pricing for adding more Dots. We flag those as not published rather than guess, and will update this guide as official details land.

This is an independent summary of OpenAI’s announcement, not a reproduction of it. Verify important details against OpenAI’s own documentation, linked at the end of every page.