OpenAI Starts Dots at $100 as AI Agent Pricing Takes Shape

ai agents artificial intelligence chatgpt dots meta muse openai recurring revenue subscription pricing subscription strategy usage-based pricing Sep 30, 2026

Meta starts Muse free and charges for more capacity. OpenAI starts individual Dot access at $100, with more AI capacity and speed higher up the ladder.

OpenAI has launched Dots, persistent AI agents that can keep working between conversations and use connected apps to get things done. The launch also shows how OpenAI is beginning to price persistent agents: individual access starts well above its lower-cost ChatGPT plans.

Dots are not included with ChatGPT Free, the $8-a-month Go plan or the $20-a-month Plus plan. Individual access begins with Pro 100 at $100 a month, with Dots also included in the $200 Pro 200 and $500 Pro 500 plans. Business Premium customers also get access, while Enterprise is currently in beta.

Meta is taking a very different approach with Muse. It launched its personal AI agent on September 8 with free access subject to usage limits, then added paid plans for people who want more capacity.

So we now have two large AI companies bringing persistent personal agents into the market with very different starting points for access and pricing.

What Is a Dot?

A Dot is an ongoing assistant that can keep context and work between conversations.

OpenAI says Dots learn a user’s preferences and goals, have their own cloud computer and browser, and can connect to more than 4,000 apps through its plugin ecosystem. A user can talk to the same Dot through ChatGPT, Slack or Microsoft Teams.

A Dot can also keep track of ongoing responsibilities and continue working between conversations. It can monitor connected information, pick work back up later and bring the user back in when it needs a decision or approval.

OpenAI gives examples such as tracking customer feedback, updating research, preparing invoices and drafting content. Users control which apps a Dot can access and can require approval before certain actions.

OpenAI is starting with one primary Dot per eligible user. The company says customers will eventually be able to add more Dots and have multiple agents working at the same time.

OpenAI Starts Dots at $100, Then Adds More AI Capacity

OpenAI’s current pricing puts a clear line between people who can use a Dot and those who cannot.

Plan	U.S. price	Dots at launch
Free	$0	No
Go	$8/month	No
Plus	$20/month	No
Pro 100	$100/month	Yes
Pro 200	$200/month	Yes
Pro 500	$500/month	Yes
Business Standard	$25/month or $20 annually per user	No
Business Premium	$125/month or $100 annually per user	Yes
Enterprise	Custom	Beta

That makes $100 a month the current entry point for an individual who wants a Dot.

OpenAI describes the first Dot as included with an eligible subscription. It is not selling a Dot separately for $100. The subscription gets the customer into the part of the product lineup where Dots are available.

Higher Pro tiers then provide more overall AI usage and, at the top end, faster processing.

Pro 500 includes OpenAI’s highest Pro usage allowance, equal to 25 times the ChatGPT Plus allowance, along with access to GPT-6 Astra Ultrafast. Ultrafast provides up to eight times faster token generation in Codex and is also available in ChatGPT Work.

Those benefits are part of the broader Pro subscription. OpenAI has not yet published permanent Dot-specific usage limits for each plan.

During the first month of the rollout, OpenAI is giving Dots extended usage treatment. Conversations with a Dot do not count against normal ChatGPT limits, while Work or Codex tasks started by a Dot use those products’ existing allowances. OpenAI says permanent Dot usage terms will come later.

OpenAI has already said customers will eventually be able to add more Dots and increase either the speed of a Dot or the amount of work it can take on each month.

That offers an early look at how agent pricing could develop. Access to the agent can be one part of the offer. How much work the customer can run through it, and how quickly that work gets done, can become separate things to package and price.

Meta Starts With Free Access

Meta’s Muse offers a useful comparison because the underlying product idea is similar, but the commercial approach starts in a different place.

Muse is also a persistent personal agent. It has its own cloud computer, remembers information about the user, works with connected services and can continue working after the app closes. Meta says Muse can handle tasks such as sending emails, booking travel, filling out forms and making purchases with approval.

Meta gives people access to Muse for free, subject to a usage limit. Customers who want more capacity can move into paid subscriptions.

Muse plan	Price	Weekly allowance
Free	$0	Usage limit
Power	$20/month	500 million Muse tokens/week
Maximum	$100/month	3 billion Muse tokens/week

Meta says the paid subscriptions are in limited testing and availability may vary.

The two companies are putting the paywall in different places.

Meta lets people start with the agent for free, then charges for more capacity. OpenAI currently requires an individual customer to reach the $100 Pro tier before Dots become available, with higher-priced plans providing more overall AI usage and speed.

OpenAI says Dots will eventually reach more users, so today’s $100 entry point may not be permanent.

Insider Take

The most interesting part of this launch for subscription operators is how quickly the product, pricing and customer behavior around AI agents are developing at the same time.

Meta launched Muse on September 8. Roughly a week later, more than 500,000 people had tried it, according to internal data reported by The Information. Three weeks after launch, Meta was already expanding Muse into small-business workflows.

OpenAI introduced Dots just 21 days after Muse, with the agent already tied to a paid subscription ladder.

These companies are not waiting for agent products to settle into a mature category before deciding how to package them. The products are changing, pricing models are taking shape and early usage is growing at the same time.

AI agents are reaching users quickly, while the products, pricing and business models around them are being built in real time.

Related Member Resources

Dots and Muse are already being packaged around access, usage and capacity. These resources help operators put the AI developments in context and work through the pricing decisions that come with products where heavier use can mean more value and more cost.

AI for Subscription Operators: Agentic Commerce, Subscription Optimization & AI in the Tech Stack
This monthly briefing looks at the AI developments most relevant to subscription businesses, including agentic commerce, subscription optimization and changes in the technology stack. It helps operators see where products such as Dots and Muse fit into the broader shift toward agents that can act for customers and businesses.
How Subscription Pricing Works: A Beginner’s Guide (Includes Worksheet)
This guide breaks a pricing decision into the offer, customer value, business economics and the way the subscription is structured. For AI products, the worksheet can help teams think through where access belongs, how much usage to include and what should change as customers move into higher tiers.