ChatGPT Dots Explained

ChatGPT dots are always-on AI agents powered by GPT-6 Astra that maintain continuous progress on user projects between conversations.

Announced on September 29, 2026, each dot operates on its own dedicated cloud computer and browser environment while connecting to external tools through plugins.

The older dot icons in the ChatGPT interface are unrelated to these agents, as explained in the guide to ChatGPT dot indicators.

ChatGPT Dots Architecture and Core Functions

A dot is an always-on ChatGPT agent on GPT-6 Astra with its own cloud computer and browser that keeps working on your goals between chats, with the first dot included in Pro and Business Premium plans.

According to the official introducing dots announcement, dots connect to over 4,000 apps through the OpenAI plugin ecosystem and operate 24/7 toward user-defined targets. Each dot learns from user feedback over time and can handle multiple projects simultaneously without forcing users to manage disconnected conversation threads.

The OpenAI Help Center article on getting started with your dot notes that a dot differs from a standard chat session by taking ongoing responsibility for work and retaining context across sessions. Users reach their dot through desktop, web, and mobile apps, as well as Slack and Microsoft Teams, while context carries across all supported channels.

What a Dot Can Do

OpenAI's launch examples show a dot taking a project and running with it while it works on several others.

OpenAI's announcement gives examples. A developer's dot watches customer feedback, builds and tests fixes, and brings complete PRs with videos. A launch lead's dot revises launch materials when scope changes, and a scientist's dot reruns analyses as data arrives. In one early tester anecdote, a dot flagged an unbilled publication, prepared an invoice, and transmitted it after receiving explicit approval.

When you aren't working with your dot, it conducts proactive research through connected apps using read-only tools that cannot send messages, change app content, or control your browser or computer. Users can also configure scheduled tasks, such as automated morning calendar reviews, or speak directly to their dot via voice calls.

• Building and testing fixes from customer feedback, then bringing complete PRs
• Continuous proposal revisions and automated proof of concept construction
• Scheduled tasks, such as a calendar review each morning, managed under Scheduled in the dot's profile
• Proactive background research using strictly read-only application permissions
• Voice call interactions and team channel messaging inside Slack and Microsoft Teams

Account Eligibility and Regional Availability

Access to dots rolls out gradually across selected paid tiers starting on September 29, 2026, with initial setup restricted to desktop environments.

Individual subscribers on Pro plans receive dot access in supported countries, excluding accounts located in the EEA, Switzerland, and the United Kingdom. In contrast, ChatGPT Business accounts with Premium seats receive dot access across all supported global regions without geographic carve-outs.

Enterprise, Education, and Healthcare workspaces can evaluate dots through an optional beta, though workspace administrators must enable the feature because access remains turned off by default. Free, Go, and Plus are not listed as including a dot, and complete setup instructions appear in the guide to setting up ChatGPT dots.

Plan Pricing and Usage Structure for ChatGPT Dots

The first dot is included without added subscription fees on eligible Pro and Business Premium accounts.

Individual Pro tiers documented on the ChatGPT Pro tiers support page include Pro 100 at $100 per month, Pro 200 at $200 per month, and Pro 500 at $500 per month. For team workspaces, the ChatGPT Business seat billing page lists Standard seats at $25 per user monthly ($20 billed annually) and Premium seats at $125 per user monthly ($100 billed annually), with Business workspaces requiring at least two users.

For the next month, dots usage will not count toward plan allowances for eligible Pro, Business, and Enterprise users. Direct conversations with a dot do not consume standard ChatGPT usage limits, though delegating tasks into Codex or ChatGPT Work draws against regular quota allowances. Detailed tier limits are outlined in the guide to ChatGPT dots pricing.

Prices change over time, so verify current figures directly on the OpenAI pricing page.

Safety Controls and Execution Boundaries

OpenAI isolates each dot on a dedicated cloud computer so your computer and its contents stay separate unless you choose to connect it.

Local device access is completely optional, requires the desktop application, and defaults to turned off until granted by the user. Dots use saved credentials on supported websites without exposing raw passwords to underlying models, while an automated review mechanism cross-checks actions against user guidelines.

Users define boundaries using Custom Rules across four execution modes: take action without asking, take action if pre-approved in prompts, ask before taking action, or hand off entirely. Sensitive administrative tasks, including account password adjustments, always require direct human execution as detailed in the guide on whether ChatGPT dots are safe.

• Custom Rules covering four granular levels of autonomous action
• Sensitive tasks such as changing a password always stay with you
• Activity View showing progress, including background work, with the option to redirect
• Default exclusion of Business, Enterprise, and Education workspace data from model training
• Complete dot resets to purge saved memories and application data

Specialist Dots and Enterprise Operations

Specialist dots operate as dedicated organizational coworkers assigned specific operational responsibilities rather than personal assistant duties.

Unlike personal dots, specialist dots feature distinct corporate identities for identity management, dedicated system credentials, and IT-provisioned infrastructure integrated with core business databases. OpenAI is starting with focused enterprise pilots, built on early testing inside OpenAI across procurement, invoice processing, email marketing, customer support and commercial contracting.

OpenAI is collaborating with Microsoft to integrate specialist dots into administrative security frameworks within Microsoft Agent 365, with comprehensive deployment patterns covered in the guide to ChatGPT dots for business.

Unpublished Details and Current Operational Limits

OpenAI has left several technical specifications and long-term commercial terms undisclosed at launch.

VentureBeat reported that OpenAI has not disclosed the numeric allowance for deeper work, additional dot pricing, or terms after the first month after launch. OpenAI has also not published the price of more speed or capacity, or specialist-dot pricing.

Because dots launched on September 29, 2026, independent third-party performance benchmarks and real-world team reliability reports are not yet available. OpenAI has stated that dots can still make mistakes and advises human verification for consequential work.

Comparison of ChatGPT Dots to Alternative Agents

The rollout of dots introduces always-on agent execution directly into ChatGPT workspaces alongside competing autonomous tools.

TechCrunch described dots as featuring cartoonish avatars and cited Meta Muse as a comparable product. VentureBeat listed Meta Muse, Instinct, Manus 2.0, Grok Bot, and OpenClaw among competing agents. Separately, Cowork merged into Claude on September 16, 2026.

A dot differs from ChatGPT Work, which OpenAI describes as an agent for longer, multi-step work and finished deliverables, as detailed in the comparison of ChatGPT dots and alternatives.

Integration Strategy for Workspace Automation

Deploying always-on agents into business operations requires establishing clear boundaries before connecting core systems.

At Layer3Labs, we build custom workflow automation and business process integrations for teams, and the primary failure mode in autonomous rollouts is granting broad execution permissions before verifying how an agent handles edge cases. Teams evaluating dots should keep local computer access turned off, mandate approval on outbound communication, and connect only the apps a first task needs.

Organizations that handle proprietary datasets or regulated compliance pipelines should maintain strict Custom Rules and route all sensitive file interactions through manual confirmation gates.

Review each app's permissions before you connect it to your dot, as OpenAI's getting started article advises.