OpenAI Dots: Always-On Personal Agents, Explained

OpenAI Dots: Always-On Personal Agents, Explained

At DevDay in San Francisco on 29 September 2026, OpenAI unveiled Dots, its take on the personal AI agent. Built into ChatGPT and powered by the GPT-6 Astra model, Dots are designed to run continuously in the background rather than wait for you to open a chat.

What a dot is

A dot is an always-on agent with its own cloud computer. Instead of being tied to one app, you reach a dot from ChatGPT, text messages, Slack, Teams, or a phone call, and it pursues goals you set with minimal oversight. OpenAI said users will eventually run several dots at once, each owning a different responsibility.

What dots can do

Dots connect to more than 4,000 applications, including workplace tools like Slack and Teams, and act inside them. In the launch demos, a dot named Alfred ended a phone call by announcing that a user's website had gone live with the design already approved, and a dot named Felipe edited a company dashboard by voice and proactively flagged a clash between a finance meeting and a family recital, offering to move it. The emphasis is on initiative: a dot that notices and acts, not just one that responds.

Pricing and availability

The first dot is included at no extra cost in OpenAI's Pro and Business Premium plans in eligible markets, with the Pro plan at $100 per month. That positions Dots as a premium, work-centric offering.

How it compares

Dots are OpenAI's answer to Meta's Muse, xAI's Grok Bot and the startup Instinct. Dots stand out for living inside the tools where work already happens. See The 2026 Personal AI Agent Wars for the head-to-head.

Explore more

Read the Dots wiki article and explore agent and workflow prompts.