techcoff.com

ChatGPT Dots: What They Are and How to Use OpenAI’s Always-On Agents

For two years, ChatGPT has waited for you to type. That era just ended. On September 29, 2026, OpenAI introduced Dots — AI agents that keep working after you close the chat and come back only when a decision needs you.

Here is what Dots are, who can use one, how to set one up, and where the limits are — including the brand-new Wallet that lets your Dot pay.

What Are ChatGPT Dots?

A Dot is an always-on AI agent. Instead of opening a new conversation for every task, you give it a standing job and it keeps working in the background.

Each Dot runs on its own cloud computer with its own Chrome browser, powered by GPT-6 Astra, and can connect to more than 4,000 apps through OpenAI’s plugin ecosystem. You can open its computer at any time to inspect what it is doing.

The key difference from a normal chat: a Dot remembers. It carries memories from ChatGPT and builds its own from your work, so the more you work together, the more it learns your preferences. You reach it in the ChatGPT app, through Slack and Microsoft Teams — or on a voice call.

When nobody has given it a task, a Dot looks for ways to help in the background — OpenAI calls this proactive research. In this mode its tools are read-only, so it can only surface things worth your attention.

What Can a Dot Actually Do?

A Dot can research a topic, draft the document, and bring the finished work back for your review.

Here is a concrete example. Say you run a small online store in Austin, Texas. You give your Dot one standing job: “Every morning, check my five best-selling products against our three main competitors and flag any price change.” While you sleep, the Dot checks the sites and prepares a short summary — and if a competitor undercut you by 15%, it flags that first and suggests a new price. You never typed a prompt. The work just happened.

Dots also handle tracking a project, preparing a trip, and monitoring an inbox. You define the goal once, the Dot does the repetitive work, and you decide the moments that matter.

How to Set Up and Use ChatGPT Dots

Step 1: Check that you can get one. Dots are rolling out to ChatGPT Pro and Business Premium users, and to Enterprise, Edu, and Healthcare workspaces. Your first Dot is included in Pro and Business Premium. Pro access excludes the EEA, Switzerland, and the UK, and you must be 18 or older.

Step 2: Create your Dot and give it a standing job. Open ChatGPT on desktop and create a Dot from the Dots section (its default handle is @yourname-dot). Describe the goal the way you would brief a new assistant: “Track our competitors’ prices every morning” works; “Do research” does not. Be specific about what success looks like.

Step 3: Connect the apps it needs. A Dot works through the apps you connect in ChatGPT’s Plugins tab. Connect only what the job actually needs.

Step 4: Set your Custom Rules. For each type of action — sharing, buying, accessing something — pick one of four behaviors: act without asking, act if pre-approved, ask first, or hand off to you. Set the sensitive ones to “ask first” on day one.

Step 5: Add custom instructions, then let it run. Since November 12, 2026, Dots can follow custom instructions (up to 35,000 characters) set in the thread or in Customize > Behavior. Use them to teach tone and format, then check the Dot’s work early — it learns from your feedback.

ChatGPT Wallet: Your Dot Can Now Pay

On November 13, 2026, OpenAI launched ChatGPT Wallet. Until now, a Dot could find a product, compare prices, and prepare a booking — then stop and hand you the checkout. With Wallet, it can complete the purchase.

You add a card to the Wallet once. When you ask your Dot to buy something, it finds the item, and before any money moves, you review the items, the total, and the card, then approve with Face ID or a fingerprint scan. Every purchase needs that approval; there is no silent spending at launch.

The Dot never sees your actual card details — the payment network issues a token for each transaction instead. The wallet was built with American Express, Mastercard, and Visa; eligible US-issued cards (including Chase) work with it, and cardholders keep their rewards and dispute protections.

The limits are real. At launch, Wallet is rolling out gradually to a limited number of US ChatGPT Pro subscribers aged 18 or older with a US billing address and an eligible US-issued card. It works only inside Dots, not ordinary ChatGPT conversations, and some websites block agent purchases, so a checkout can still fail at the merchant’s end. Pre-approved purchases are a future step, not a shipped feature.

Your Dot Can Now Text You

On November 13, 2026, OpenAI started rolling out texting for Dots as a preview for Pro users in the United States — your Dot can message you through RCS in Google Messages, or iMessage on iPhone.

Setup is simple: open the Dot’s profile in the ChatGPT app and select Texting. Texting is one-to-one only — no group chats, no phone calls — and approvals happen through a link back into ChatGPT. One privacy note: OpenAI uses Linq, a third-party messaging provider, for the texts.

What Dots Can’t Do Yet

Dots launched with real gaps. No password changes — built-in stops cover passwords, money, deletes, and installs. No silent spending — every purchase needs your approval and a biometric check. No email of its own — a Dot has no email address, cannot call you, and you cannot view or edit its memories; reset or delete the Dot to clear them. No API — developers reach Dots through plugins and MCP Events. Regional limits — Pro access excludes the EEA, Switzerland, and the UK, and texting is US-only. OpenAI’s own tests warn that a Dot juggling too many unrelated tasks is more likely to step outside its instructions. The honest summary: a Dot is a capable junior colleague, not an autopilot.

Who Can Use Dots and What It Costs

ChatGPT Pro and Business Premium users in eligible markets get their first Dot included at no extra cost. For Enterprise, Edu, and Healthcare workspaces, Dots became generally available on November 4, 2026, with an admin preview until November 18 — free to try during the preview, billing from November 18, and on by default after that date unless an admin opts out. New Dots also support data residency and Enterprise Key Management, and OpenAI says Dots are HIPAA-eligible in covered workspaces.

Since early November, one Dot can keep several of your local computers connected and choose which runs each task.

If you run a small business, a Dot is worth trying the moment you have a recurring job that eats an hour a week. Our roundup of free AI tools for small business marketing in 2026 pairs well with Dots, and these ChatGPT prompt ideas for content creation will help you write instructions a Dot can actually follow.

Frequently Asked Questions

What are ChatGPT Dots?

Dots are always-on AI agents inside ChatGPT. Each Dot runs on its own cloud computer, connects to your apps, remembers context, and keeps working toward goals you set — even when you are not in the chat. They are powered by GPT-6 Astra, announced at OpenAI’s DevDay on September 29, 2026.

How much do ChatGPT Dots cost?

Your first Dot is included in ChatGPT Pro and Business Premium at no extra cost. For Enterprise, Edu, and Healthcare workspaces, Dots went generally available on November 4, 2026, with a free preview until November 18. Work delegated to ChatGPT Work or Codex is billed separately.

Can a Dot buy things for me?

Yes, with ChatGPT Wallet, launched November 13, 2026 for US ChatGPT Pro subscribers aged 18 and older with a US billing address and an eligible US-issued card. The Dot never sees your real card details — the network issues a token per transaction — and every purchase needs your approval with a Face ID or fingerprint check.

Can my Dot text me?

Yes, as a preview for Pro users in the United States, announced November 13, 2026. Your Dot can send updates through RCS in Google Messages on Android or iMessage on iPhone. Texting is one-to-one only, and approvals happen through a link back into ChatGPT.

The Bottom Line

ChatGPT used to be a tool you opened. Dots turn it into a colleague you brief. Give one a clear standing job, tight permissions, and a week of feedback, and it starts returning finished work instead of waiting for your next prompt. Wallet and texting are still rolling out, and the limits are real — but the direction is obvious. The people who learn to manage an agent now will be the ones whose work keeps happening while everyone else is still typing.

Leave a Comment