OpenAI Dots Explained: What ChatGPT Dots Are and Cost (2026)

OpenAI announced dots on September 29, 2026 at DevDay in San Francisco. A dot is a named, avatar-style agent that lives inside ChatGPT, takes ongoing responsibility for work you hand it, and keeps making progress when you are not talking to it. It can research in the background, run scheduled jobs like calendar reviews, message you in Slack or Microsoft Teams when it needs a decision, and hand coding work to Codex. The rollout is gradual, so being on an eligible plan does not mean the feature has reached your account yet.

Dots matter because they are OpenAI's direct answer to two agents that shipped weeks earlier: Meta's Muse personal agent (September 8) and SpaceXAI's Grok Bot (beta since August 11). This guide covers what ChatGPT dots are, who can get them, what they cost, how to set one up, the safety controls, the known limits at launch, and how dots compare with Muse and Grok Bot. Every fact below is sourced; where reports disagree, we say so.

What are ChatGPT dots?

OpenAI describes a dot as an always-on agent inside ChatGPT that takes ongoing responsibility for work and keeps making progress between conversations. In practice that breaks down into a few concrete pieces:

• Model: dots run on GPT-6 Astra, OpenAI's current frontier model.
• Its own computer: each dot gets a dedicated cloud computer and browser, so it can keep working when your laptop is closed. Connecting your own machine is optional.
• Apps: dots connect to 4,000+ apps through plugins, plus Slack and Microsoft Teams.
• Memory: a dot inherits your existing ChatGPT memories and builds new ones from activity in the apps you connect.
• Background work: when idle, a dot looks for ways to help using connected apps with read-only tools.
• Personality: you name and customize your primary dot. TechCrunch describes the avatar as "bubbly" and cartoonish, which is where the name comes from.

OpenAI's own examples lean practical: a developer who has a dot monitor customer feedback and implement fixes, a scientist who has it rerun analysis on new experimental data. TNW quotes an early tester whose dot noticed an unsent freelance invoice in an email thread, pulled the details, and drafted it. Latent Space's DevDay recap mentions testers who had their dot negotiate with customer service to cut roughly $500 a year in charges. Those are anecdotes, not measured results.

Is a "dot" the same thing as a ChatGPT agent?

It is the next step. Earlier ChatGPT agent features ran a task when you asked and stopped. A dot is persistent: it keeps a goal, a computer, a memory, and connected apps, and it reports back on its own schedule. OpenAI says teams of dots working together is the longer-term plan; at launch you get one primary dot. Enterprise pilots can also run "specialist dots" with their own identities, credentials, and tools, with direct support from OpenAI engineering.

Who can get OpenAI dots?

Eligibility depends on your plan, your region, and your age. The table below summarizes what OpenAI and launch coverage report.

Two caveats. First, the rollout is gradual and can take several days to reach an eligible account. Second, the European exclusion applies to Pro only; Fortune ties it to stricter data privacy law. If you are a Pro user in London or Berlin, you will not see dots yet, while a Business Premium workspace in the same city can.

How much do ChatGPT dots cost?

There is no separate price for your first dot. It is included in ChatGPT Pro and Business Premium. For what those plans themselves cost after the DevDay changes, see our breakdown of the new ChatGPT Pro and $500 plan.

What counts against your usage is less settled, and sources describe it differently:

• Chatting with your dot does not count toward ChatGPT usage limits. All sources agree on this.
• Background work: DataCamp and The Decoder (Mixed) say dot usage draws on a plan allowance, but that it will not be counted for the first month after launch, and OpenAI will publish per-tier terms after that. Latent Space's recap says primary dot work does not draw on plan usage at all. Treat the month-one grace period as the confirmed part and the long-term terms as unpublished.
• Codex and ChatGPT Work tasks a dot spins up count against your normal limits like any other Codex task.
• Extra dots: OpenAI says you will later be able to pay a flat monthly fee to add more dots or raise a dot's speed and monthly work allowance. It has not named a price.

So the honest answer today is: one dot is free on the top plans, heavy use is free for the first month, and nobody outside OpenAI knows what month two looks like.

How do you set up a dot in ChatGPT?

Setup only works on desktop. You can message a dot from mobile afterwards, but you cannot create or configure one on a phone.

1. Open the ChatGPT desktop app, or ChatGPT in a desktop web browser, on an eligible Pro or Business Premium account.
2. Create your dot and follow the onboarding prompts. You can name and customize it.
3. Connect the apps you want it to use (email, calendar, Slack, Teams, and so on from the plugin catalog).
4. Let it introduce itself, then give it its first ongoing goal.
5. Set boundaries: go to Settings > Personalization > Permissions > Custom rules, select Add, describe the action, choose how the dot should handle it, and select Add rule.

Enterprise, Edu, and Healthcare users need an admin to enable the beta for the workspace first.

Where can you talk to your dot?

• ChatGPT on desktop, web, and mobile (text and voice calls inside ChatGPT)
• Slack and Microsoft Teams
• Texting: a limited beta for Pro users in the US only

A dot can message you with progress, questions, or decisions it needs, but at launch it cannot place a call to you and does not have its own email address.

What safety controls do dots have?

An agent that holds your logins and works while you sleep is only as good as its brakes. OpenAI's controls stack in layers.

Custom Rules: four levels of autonomy

Every rule you add maps an action to one of four behaviors:

DataCamp points out that "pre-approved" is narrower than it sounds: it means you asked for that exact action in your prompt, not that you approved a category in advance. Custom Rules can add restrictions (for example "never send emails") but cannot override OpenAI's built-in guardrails.

Built-in guardrails

• Read-only background mode: when a dot works proactively, its tools are restricted to read-only. It cannot send messages, change app content, or control your browser or computer in that mode.
• Auto-review: a separate system checks actions that could affect your accounts or share information against your instructions and safety requirements before they run. It cannot be turned off.
• Mandatory approval: sensitive tasks such as changing a password, permanently deleting data, or installing software always stay with you.
• Monitoring: OpenAI says monitoring can pause or stop a dot if it detects a safety concern.
• Credentials: Engadget reports saved passwords are stored without being exposed to the underlying model.

What critics are saying

The main criticism is that all of this is self-reported. OpenAI has published no benchmarks, success rates, or independent evaluations of dots, and says itself that a dot "can make mistakes, including when following your rules." Commentators want independent security testing of an auto-approved agent on tasks designed to escape the read-only boundary before trusting it with real accounts. Fortune notes that, like Muse, dots collect substantial personal data by learning preferences and remembering conversations.

Timing did not help. On September 25, four days before launch, OpenAI disclosed that some of its internal research agents (not dots) had posted 53 ChatGPT users' images to public image hosts. It is a separate system, but it is exactly the class of failure people worry about with always-on agents. For the wider picture on the model underneath, see our write-up on GPT-6 Astra's safety and cyber capabilities.

What are the limits of dots at launch?

• One primary dot per user; teams of dots are planned, not shipped.
• Setup and configuration are desktop-only.
• A dot cannot call you and has no standalone email address.
• You cannot delete a dot's own saved memories without deleting the dot. Disconnecting an app does not erase what the dot already learned from it.
• Texting is a US-only Pro beta.
• No Pro access in the EEA, Switzerland, or UK.
• Long-term usage terms and extra-dot pricing are unpublished.

The memory limit is the one to think about before connecting a sensitive inbox. If you want a clean slate later, the only path is deleting the dot and starting again.

OpenAI dots vs Meta Muse vs Grok Bot: which is better?

All three are "agent with its own cloud computer" products that act inside your signed-in apps and pause for approval on risky steps. They differ on price, distribution, and who they are built for.

Short version:

• Pick dots if you already pay for ChatGPT Pro or Business Premium and your work lives in Slack, Teams, and the apps in OpenAI's plugin catalog. The Codex hand-off is the strongest developer angle.
• Pick Muse if you want the cheapest entry and live in WhatsApp. It is the only one with a free tier. Our Meta Muse agent guide covers it in detail.
• Pick Grok Bot if your team already pays for Cursor or SuperGrok and wants an agent for business workflows next to the IDE. See our Grok Bot guide.

What was the dot.com Grok Bot stunt?

On launch day, an X user noticed that dot.com redirects to the download page for xAI's (now SpaceXAI's) Grok app. TechCrunch notes that Whois records show the domain was transferred in July 2026, before OpenAI announced the product name. The internet read it as trolling. TechCrunch notes it is "entirely possible" xAI bought it for normal reasons, possibly to catch typos of "bot," though xAI does not own bot.com. Funny, but it tells you nothing about which agent is better.

Who should turn on a dot?

• Solo founders and freelancers on Pro: yes, with tight Custom Rules. Start with "ask before taking action" on anything that sends messages, and let it earn autonomy on drafting and research.
• Engineering teams: useful for bug triage, watching failing builds, and routing fixes to Codex. Remember Codex tasks it spawns still count against your limits. If you are also evaluating the newest model, read our GPT-6.1 Sol guide.
• Enterprises: wait for your admin, and use the beta period to write rules. Specialist dots, with Microsoft Agent 365 integration in development, are the enterprise story, but they are pilot-only today.
• Anyone in regulated work: hold off on connecting sensitive inboxes until OpenAI publishes long-term usage terms and some independent testing exists. The memory-deletion limit makes it hard to undo.
• Pro users in the EEA, Switzerland, or UK: not available yet. Muse is also US-first, so Grok Bot or a Business Premium workspace are the realistic options.

FAQ

What are OpenAI dots?

Dots are always-on personal agents inside ChatGPT, announced at DevDay on September 29, 2026. Each runs on GPT-6 Astra with its own cloud computer and browser, connects to 4,000+ apps, remembers context, and keeps working toward goals between conversations.

Are ChatGPT dots free?

Not on free ChatGPT. The first dot is included at no extra cost with ChatGPT Pro and Business Premium. Chatting with a dot does not count toward limits, background usage is not counted for the first month, and pricing for extra dots has not been announced.

Can I use dots in the UK or EU?

Not on ChatGPT Pro at launch: Pro access excludes the EEA, Switzerland, and the UK. Business Premium workspaces get dots in all supported ChatGPT regions, and Enterprise, Edu, and Healthcare workspaces can enable a beta.

Can I set up a dot on my phone?

No. You create and configure a dot in the ChatGPT desktop app or ChatGPT on a desktop browser. After setup you can message it from mobile, Slack, or Teams.

Can a dot change my passwords or spend money on its own?

OpenAI says sensitive tasks such as changing a password, permanently deleting data, or installing software always require your approval, and an auto-review system checks account-affecting actions before they run. You can add stricter Custom Rules, such as handing payments off to you.

How are dots different from Meta Muse and Grok Bot?

All three run on dedicated cloud computers and pause for approval on risky actions. Dots are bundled with top ChatGPT plans and live in ChatGPT, Slack, and Teams. Muse has a free tier and $20/$100 plans and lives in its app and WhatsApp. Grok Bot is included with any paid Cursor plan or a linked SuperGrok or X Premium+ subscription.

What model powers ChatGPT dots?

GPT-6 Astra, according to OpenAI's launch materials and coverage from TechCrunch, DataCamp, and TNW.

Sources

• Cursor Help: Grok Bot plans and billing
• OpenAI: Introducing dots
• OpenAI Help Center: Getting started with your dot
• TechCrunch: OpenAI launches dots, its bubbly agentic avatar
• DataCamp: OpenAI dots
• TNW: OpenAI launches dots, always-on AI agents with their own cloud computers
• Latent Space: AINews on OpenAI DevDay 2026
• Engadget: Dots are OpenAI's new personal agents
• The Decoder / Mixed: Dots exclude Pro users in the EEA, Switzerland and the UK
• Fortune: OpenAI takes on Meta's Muse with dots
• TechCrunch: The internet is convinced xAI trolled OpenAI's dots launch
• TNW: OpenAI says its agents posted 53 ChatGPT users' images online
• Neomanex: Meta Muse is live at $20 and $100 a month
• InfoQ: SpaceXAI launches Grok Bot
• Memeburn: SpaceXAI Grok Bot is live

If your team is building products on top of agents like these, Codersera can help you hire vetted remote developers.