Has AEQI stopped being for people starting a company?

No. Founding is still one of the three doors, and it is the one AEQI was built through — this company runs on its own product. What changed in August 2026 was that it stopped being the only sentence on the front page. The promise moved to running a Company, which is equally true whether the Company is three days old or thirty years old.

Three groups, and in August 2026 the product opened with a door for each. One person who wants their own work to run without them holding every thread. A founding team building something new that needs functions filled before it can afford to fill them with people. And an established company that already has customers, revenue, and operations, and wants agents inside those operations rather than a chat window beside them.

How is the AI paid for?

In credits, at one rate everywhere: $20 buys 1,000 credits, on every plan and every top-up. Every size includes a monthly amount, and it pays for the work your agents actually do — model calls and tool use — on a machine provisioned for your Company alone. It is a meter of work performed, not of people added, which is why the price still makes sense for a company that grows output while headcount stays flat.

From starting a company to running one

In August we changed what AEQI is for: from starting a company to running the one you have. What changed, and for whom.

August 22, 2026 · Luca Eich

AEQI positioning: from starting a company to running one — three groups, one company.

Written August 22, 2026.

AEQI has changed since: today it starts a company that runs itself, with eleven AI teammates that build your website, find leads and write your emails. You approve every move. The pricing below is today’s; the rest stands as written. See AEQI today

For most of its life, the first sentence on this site was about founding: start a company that runs itself. It was a true sentence and a good one, and it had a ceiling we could see from the beginning. It addressed the moment before a company exists — a moment most of our buyers had already passed years ago.

The sentence became “Make your Company run itself.” That is a small edit and a large change of position, and it is worth explaining rather than quietly shipping.

What the pilots taught us

The companies that came to us were not pre-formation. They were law firms with more instructions than associates. Agencies whose founder was the bottleneck on every deliverable. Mid-sized operators with a compliance officer, a real ledger, and a backlog of work nobody had time to do. Not one of them wanted to start anything. They wanted the thing they already had to stop depending on them for every step.

Founding remained the door we knew best — AEQI is built and run by agents inside its own product, and that is still the sharpest proof we have. But the proof was being used to argue for the wrong moment. A company that already exists does not need a founding ceremony. It needs staffing.

The three groups

So the product opened with three doors rather than one, and the rule we hold ourselves to is that every word on the front page must be equally true for all three.

One person, running their own work. A freelancer, a consultant, a solo operator. The unit of the problem is not a company in the legal sense — it is everything that only happens because they personally do it. One agent, one seat, and a place for the operating context to live outside their head. The measure of success is whether a week away costs anything.

A founding team, building something new. Two or three people who need marketing, support, finance, and engineering to exist before they can afford to hire any of them. This is the case AEQI was born in: the roles get filled by agents, the work gets decomposed into quests, the website is live from the first week, and the operating history starts compounding before there is a company to compound it in.

An established company, adding capacity. Customers, revenue, staff, processes that work. What they want is not a new company but agents inside the one they have: sitting in named roles, holding real permissions, doing work that shows up in a ledger their people can read. The demand is not replace my team. It is give my team a team.

The same thing, for one or for all

The temptation with three audiences is to build three products. We deliberately did not. Every door leads to the same object — a hosted Company with agents in roles, a quest ledger, compounding memory, a live website, and its own server. What differs is who is on the computer.

That is why there is one price. Start is $120 a month with 5,000 credits a month; Grow ($300) and Scale ($750) are the same Company on a bigger computer, with 12,500 credits and 31,250 credits a month. Every company starts with 7 days free, and paying yearly gets you 12 months for the price of 11. Above the sizes, Enterprise stays a conversation rather than a card.

The door you enter through never changes what you pay. A sole trader who has been in business for a decade is right to buy Start, and forcing them upward would be a way of losing a customer while congratulating ourselves on positioning. The size of the computer and the AI each month are the only things that move between sizes. Capability does not.

What the AI is counted in

Credits, at one rate everywhere: $20 buys 1,000 credits, whatever plan you are on and whatever you top up with. Every size includes a monthly amount, and it pays for the work the agents do on a machine that carries your database, your files, your site, and your integrations — a machine that exists whether or not you use it that week.

We are specific about the rate because the market is full of credits that mean something different on every page. Ours convert to a dollar figure at a single fixed rate, so you can compare it with anything. If you are comparing pages, compare the deliverable: a metered conversation, or a company that keeps running when you close the tab.

It also explains why the meter counts work rather than seats. Per-seat pricing is a bet that headcount keeps rising, and a vendor on that meter cannot honestly recommend doing more with fewer people. We would rather charge for output and let you decide how many humans stand next to it.

What has not changed

The thesis is the same one we started with: a company is a computer, and the parts of it that were held together by people remembering things can be compiled. Agents give it motion; roles give it authority; memory gives it continuity; the ledger makes all three legible. None of that was invented for the new sentence.

What changed is the moment we address. Not the day you decide to start a company — every day after it, when the thing you built still needs running.