Skip to main content
OpenAI dots logo

OpenAI dots

OpenAI's always-on AI agents with their own cloud computer

Editorially reviewedChecked Oct 2026How we review·60 visited this week
chatbots#openai#ai agents#autonomous agents#always-on
Teams Mobile app
Toolglade’s take

dots are OpenAI's entry into always-on, autonomous agents, and they land squarely against Grok Bot and Meta Muse: a persistent agent with its own cloud computer and browser that acts across your connected apps and reports back. The genuinely notable parts are the permission model (you set what it can do alone, what needs approval, and what it must never do) and the Auto-review safety layer that pauses a dot that seems to be going off-script. The cautions are real and OpenAI says so itself: the product is days old, it can still make mistakes, and prompt-injection protection is imperfect, so keep approval gates on for anything that sends messages, spends money, or touches production. Pricing is gated to Pro and Business Premium (the first dot is included), and availability is limited by region at launch. Verify current capabilities and rates directly with OpenAI.

Should you use OpenAI dots?

Best for
ChatGPT Pro / Business Premium users · People offloading long autonomous tasks · Teams already living in Slack, Teams, and Google Workspace
Not ideal for
Users outside eligible launch regions · Teams wanting a flat low-cost or free option · Anyone needing mature, audited agent controls today
Biggest strength
Persistent, always-on agent with its own cloud computer and browser
Biggest weakness
Days-old product; capabilities and limits still settling
Pricing
From $100
Top alternative
Uniphore · see all

Facts & pricing last checked October 2026. How we review

About OpenAI dots

dots are OpenAI's always-on AI agents (announced DevDay, Sept 29 2026): persistent agents with their own cloud computer and browser that work toward a goal across 4,000+ connected apps and report back. Powered by GPT-6 Astra; available to ChatGPT Pro and Business Premium.

dots are OpenAI's always-on AI agents, announced at DevDay on September 29, 2026. Unlike a chat session that ends when you close the tab, a dot is persistent: it runs on its own cloud computer and browser, keeps working toward a goal you set without constant prompting, and brings results back for you to review. OpenAI reports dots are powered by its GPT-6 Astra model, and you can reach a dot from the ChatGPT app, the web, Slack, Microsoft Teams, or a phone call, with text messaging planned. Setup follows a permission model: you give a dot a goal, connect the accounts it needs (dots can use the 4,000-plus apps that already work with ChatGPT, such as Slack and Google Drive), and set what it may do on its own, what needs your approval, and what it must never do. By default, significant actions wait for approval. For example, a dot can draft a message but should not send it to a person or another agent unless you asked it to. Reported example tasks range from booking a table to monitoring customer feedback, implementing and testing minor code fixes, and submitting pull requests with video documentation. On safety, OpenAI layered in extra monitoring: a system publicly called Auto-review (internally Guardian) that can pause or stop a dot if it appears to misinterpret an instruction, on top of existing ChatGPT and Codex safeguards. OpenAI itself cautions that dots can still make mistakes and that its prompt-injection protection is not perfect, so you should review consequential work. dots are rolling out in ChatGPT to Pro and Business Premium subscribers in eligible markets, with the first dot included at no extra cost and the Pro rollout initially excluding the EEA, Switzerland, and the UK. Because the product is days old, treat capabilities, pricing, and availability as a fast-moving snapshot and confirm current details with OpenAI.

TL;DR

dots are OpenAI's always-on AI agents (DevDay, Sept 29 2026): persistent agents with their own cloud computer and browser that pursue a goal across 4,000+ connected apps and report back. Powered by GPT-6 Astra; included for ChatGPT Pro and Business Premium.

Company overview

dots are a product of OpenAI, the company behind ChatGPT and the GPT model family, unveiled at its DevDay conference on September 29, 2026. They extend OpenAI beyond chat into autonomous agents that do work rather than only answer, competing directly with xAI's Grok Bot and Meta's Muse.

OpenAI frames dots around a permission model and an Auto-review safety layer, and distributes them through ChatGPT's Pro and Business Premium tiers rather than as a standalone product.

Product features

Each dot runs persistently on its own cloud computer and browser, works toward a goal you set, and connects to the 4,000-plus apps that already work with ChatGPT. You set what it can do alone, what needs approval, and what it must never do; significant actions require approval by default. dots are reachable from the ChatGPT app, web, Slack, Teams, and phone, with texting planned.

Reported tasks include booking, monitoring customer feedback, and implementing and testing minor code fixes with pull requests. An Auto-review (Guardian) system can pause or stop a dot that appears to misinterpret an instruction.

Target market

ChatGPT Pro and Business Premium subscribers who want to offload long, multi-step work to an autonomous agent, especially individuals and teams already working across Slack, Teams, and Google Workspace.

Buyer personas

End users

People who hand a dot a goal and review its results.

Buyers

ChatGPT Pro / Business Premium subscribers and team admins.

Key influencers

AI agent, developer, and productivity communities.

Ideal customer profile

A Pro or Business Premium user with repetitive, multi-app work who is comfortable granting scoped autonomy and keeping approval gates on sensitive actions.

Funding & performance

dots are an OpenAI product; OpenAI is heavily funded (including by Microsoft) with large reported rounds. Specific figures are not detailed here — verify via public sources.

Pros & cons

Pros

  • Persistent, always-on agent with its own cloud computer and browser
  • Connects to 4,000+ apps that already work with ChatGPT
  • Clear permission model: allow, require approval, or forbid actions
  • Auto-review safety layer pauses a dot that seems to misinterpret
  • Reachable from ChatGPT, Slack, Teams, and phone
  • First dot included in Pro or Business Premium; no extra usage charge

Cons

  • Days-old product; capabilities and limits still settling
  • OpenAI says it can still make mistakes — review consequential work
  • Prompt-injection protection is, per OpenAI, not perfect
  • Gated to higher-priced Pro and Business Premium plans
  • Pro rollout initially excludes the EEA, Switzerland, and the UK
  • An agent with broad app access warrants tight permissions

Pricing plans

Included with Pro
From $100 / month
  • First dot included in ChatGPT Pro
  • Reported Pro tiers: Pro 100 / 200 / 500 ($100/$200/$500)
  • Conversations do not count against usage limits
  • Verify current pricing with OpenAI
Business Premium
Per seat / month
  • dots included with a Business Premium seat
  • Team controls and connected apps
  • Confirm current seat pricing with OpenAI

Key features

Mobile app
Team collaboration
Multi-language
Integrations
4,000+ ChatGPT-connected apps, Slack, Microsoft Teams, Google Drive, phone call
Input types
text, natural language goals, voice (phone)
Output types
completed tasks, drafts, pull requests, reports
Best For
long-running autonomous tasks, work across connected apps, monitoring and follow-up

Compare key features

View all alternatives →
Feature
OpenAI dots
Uniphore
Claude
Pricing
Paid
Paid
Freemium
Free plan
No
No
Yes
Free trial
No
No
No
API
No
Yes
Yes
Self-hosted
No
Yes
No
Team support
Yes
Yes
Yes

Frequently asked questions

What are OpenAI dots?+

dots are OpenAI's always-on AI agents, announced at DevDay on September 29, 2026. Each dot runs on its own cloud computer and browser, works toward a goal across your connected apps, and reports back for review.

How much do dots cost?+

dots are included for ChatGPT Pro and Business Premium subscribers; the first dot is free within the plan. Reported Pro tiers run $100 to $500 per month. Confirm current pricing with OpenAI.

Are dots safe to let run on their own?+

They use a permission model (allow / require approval / forbid) and an Auto-review layer that can pause a misbehaving dot. OpenAI still cautions they can make mistakes and that prompt-injection protection is imperfect, so review consequential actions.

How are dots different from ChatGPT?+

ChatGPT answers in a session you drive. A dot is a persistent agent that keeps working on a goal on its own cloud computer and returns results, acting across connected apps with your permissions.

Reviews

Write a review

Pick a rating
Loading reviews…
Compare

Compare OpenAI dots with other AI tools

Side-by-side pages for pricing, features, and best-fit use cases.

All comparisons →

People also compare

From the blog

All articles →

Featured in these roundups

Similar tools you may like

Learn to use OpenAI dots well

Free, hands-on Toolglade courses that cover this kind of tool — start reading in a minute, no sign-up needed.