Dots Guides and Explainers

OpenAI's Dots are always-on AI agents with their own cloud computer and browser, the ability to write and test code, connections to 4,000+ apps, and a habit of learning from feedback over time. They were announced on 29 September 2026, are powered by GPT-6 Astra, and can keep several projects moving at once.

That breadth is exactly why a single page cannot cover everything. This library groups our guides the way people arrive: some visitors want to know what a Dot actually is, others already have access and need setup, money, or safety answers, and teams want to know where a Dot fits. Every page is written to be read on its own, and each one links back to the pillars and to the wider set.

Every page here follows the same editorial rule: state what OpenAI has published, link the official source, and mark the gaps clearly. Where a question has no published answer — for example the exact price of additional Dots beyond the first, or the full list of markets — the page says "not published" instead of guessing. That restraint is the point of an independent guide: it should be safe to act on what you read here, and safe to know what is still open.

This is an independent guide, not an OpenAI product. Facts are checked against OpenAI's launch announcement and help documentation; where OpenAI has not published a figure, we say so rather than estimate.

Start with the foundations

If you only read two pages on this site, make them the explainer and the model page. The first answers what a Dot is and how the always-on model of work differs from chatting with an assistant. The second covers GPT-6 Astra, the model named for the Dots launch. The FAQ collects the short-form answers in between.

What are OpenAI Dots? The plain-English explainer: what an always-on agent is and how a Dot works. GPT-6 Astra The frontier model that powers every Dot, and what that means in practice. Dots FAQ Short answers to the questions people ask most about Dots.

Getting access and understanding cost

Dots are rolling out to Pro and Business Premium in eligible markets, with the first Dot included at no extra cost. Enterprise, Edu and Healthcare workspaces can try the beta when an administrator enables it. Our setup and pricing guides walk through eligibility, the practical first steps, and the parts of the cost model OpenAI has left unpublished, including pricing for extra Dots. The channels guide covers reaching a Dot through ChatGPT, Slack and Teams, and the texting waitlist.

How to get Dots Eligibility, setup steps, and what to connect first. Dots pricing Plans, allowances, and the figures OpenAI has not published. Dots in Slack and Teams Reaching a Dot where your team already works.

Teams, roles and safeguards

When several people depend on the same agent, the questions change. Specialist Dots are organizational agents with their own identity, credentials and access to systems of record, and OpenAI is testing them in areas such as procurement, invoice processing, email marketing, customer support and commercial contracting. The use-cases hub collects the five scenarios OpenAI highlighted and links to role-specific guides. The security and privacy guide explains isolation, read-only proactive research, auto-review and Custom Rules.

Specialist Dots Organizational agents with their own identity and systems access. Dots use cases The five scenarios OpenAI highlighted, with role-by-role guides. Security and privacy Isolation, read-only research, auto-review, and Custom Rules.

Reference and tools

For quick lookups, the glossary defines the vocabulary of the Dot docs, and our free Agent Readiness Scorecard helps you judge which of your own workflows are ready for an always-on agent and which are not yet. Dots can make mistakes, so treat any consequential output as something to review before you rely on it.

Glossary The terms used across the Dot docs, defined in plain English. Agent Readiness Scorecard Our free tool for scoring which workflows are ready for an always-on agent.

What is still unpublished

A useful guide is honest about its edges. Dots are rolling out to Pro and Business Premium in eligible markets, with Enterprise, Edu and Healthcare able to try the beta when an administrator enables it. Beyond that, several details remain open: OpenAI says Dots will expand to more users "soon" but has given no date, and it has not published pricing for adding more Dots or scaling speed and monthly work, nor a full list of eligible markets. Plan allowances include room for "deeper work", with extended limits in the first month, but the specific numbers are not published either.

We revisit each page as OpenAI publishes more, and stamp every guide with the date its facts were last checked. If a figure is missing today, the relevant guide will say so — and will change once the source does. When in doubt, verify important details against OpenAI's own documentation rather than relying on any third-party summary, including this one.

How to use the library

Read the explainer so the vocabulary is clear.
Check eligibility and pricing before planning a rollout.
Review the security guide before connecting real systems.
Use the use-cases hub to pick a first workflow.
Reserve anything consequential for human review.

Source: OpenAI, Introducing dots and the Dots help center.