What an OpenAI Dot Actually Costs — and When Always-On Is Worth It
OpenAI's Dots have no free tier: one dot comes with the $100/mo ChatGPT Pro plan, and each dot runs continuously on its own cloud computer. That "always-on" design is the whole value — and the whole cost question. Here's how to decide if a Dot earns its keep.
What an OpenAI Dot Actually Costs — and When Always-On Is Worth It
OpenAI introduced Dots at DevDay on September 29, 2026 — "always-on" agents that pursue a goal in the background instead of answering once and stopping. Each dot runs on its own cloud computer with its own browser, reaches roughly 4,000 apps through ChatGPT, and is powered by GPT-6 Astra. Sam Altman called it his favorite OpenAI product.
It's genuinely new. It's also the one personal agent of 2026 with no free tier, so the first honest question isn't "what can it do" — it's "is always-on worth paying for?"
The price, plainly
There is no free Dot. At launch, one dot is included with the $100/month ChatGPT Pro plan (and with Business Premium), with Enterprise in beta. So you're not buying a dot à la carte — you're either already paying for Pro, or you're paying $100/mo largely to get one. And availability gates it further: Dots are not available in the EEA, Switzerland, or the UK at launch, so for a lot of people the cost is currently "not available at any price."
Why "always-on" is the cost, not a feature
The thing that makes a Dot special is also what makes it expensive to run well. A normal ChatGPT session costs you when you use it and stops when you close it. A dot keeps working — looping, checking, acting — against a goal you set, on its own machine, whether or not anything is happening. That persistence is the value (it catches the price drop at 2 a.m., it keeps chipping at a long task) and the liability (it burns effort on the 99% of checks where nothing changed).
So the economics of a dot come down to one question: does this goal actually benefit from being watched continuously? Two honest buckets:
- Worth a dot: goals where freshness has real value and the window is unpredictable — monitoring something that can change at any time and acting the moment it does, or grinding through a genuinely long, multi-step task that would otherwise sit unfinished.
- Not worth a dot: anything you could run once a day, or ask for on demand. "Summarize my inbox each morning" doesn't need an always-on agent — a scheduled task or a quick ChatGPT prompt does the same job without a persistent worker.
If your use case is in the second bucket, a dot is a worse deal than the ChatGPT you already have. The Pro plan's value has to come from Pro, with the dot as a bonus — not the other way around.
The honest caveats
Early users praised real wins (inbox zero, long tasks finished) but flagged two things worth knowing before you trust a dot unattended: slow speeds, and occasional unapproved actions. Dots do have a permission model and an Auto-review safety layer, but the reports are a reminder to keep write-access tight and start read-only — let a dot propose before you let it act, especially with email or anything that spends money.
The verdict
A Dot is worth it if you're already a ChatGPT Pro user (or would be anyway) and you have a goal that genuinely rewards continuous, autonomous attention. If you're weighing $100/mo mostly to get a dot for tasks you could schedule or ask for on demand, the math doesn't favor it yet — and if you're in the EEA, UK, or Switzerland, it's not an option at all for now. Prices and availability here move fast; confirm both with OpenAI before you commit.
Comparing the field? See Grok Bot vs Meta Muse vs OpenAI Dots, and the free course on running personal AI agents safely and cheaply. Full listing: OpenAI Dots.