Why a pay per minute AI twin beats a subscription for experts

This page covers experts who charge by the minute for an AI version of themselves. It doesn't cover AI phone agents or voice software billed per minute, or the digital twins used in engineering.

What is a pay per minute AI twin?

A pay per minute AI twin is an AI version of you that visitors pay to talk to by the minute, built from things you've already written and recorded. It talks to visitors in text chat, or as a live talking avatar with your cloned voice. You build it once, from a photo, a short voice clip and your material. It answers from your own material and the decision rules you approved. Every twin is labelled as AI, and it gives information, not regulated advice.

Pay per minute is how it earns. You set a price per minute above a platform base of €0.50 a minute. Visitors buy credits through Stripe and spend them while they talk to your twin. A text question costs a quarter of a minute. A live avatar session is billed per started minute, with a one-minute minimum. You keep 80% of the markup, meaning the amount above the base. Our 20% commission applies to the markup only, never to the base.

Building a twin is free and there's no subscription, so a quiet month costs you nothing beyond any free taster you choose to fund. Optional paid seats add extras, such as an allowance of free minutes for your visitors, and analytics. They're listed on the pricing page, and you don't need one to go live.

Why is a subscription an awkward fit for an expert's AI twin?

A subscription charges people every month, and questions for an expert don't arrive every month. Someone has a pressing question about their pricing in March, then nothing until a new client turns up in October. A monthly fee asks that person to guess how often they'll need you for the rest of the year. Most people won't guess, so they don't sign up.

A subscription also gives you a job. Every month has to feel worth renewing, so you end up making content to stop people cancelling. That's a lot of work that has little to do with the questions they came with.

Per-minute pricing moves the decision to the moment it makes sense. The visitor has a question now, sees your price and decides whether it's worth a few minutes. Nobody subscribes to you. Some visitors keep a monthly credit plan on the platform, and those credits get spent at whatever rate each expert sets. Credits never expire, so nobody loses money because a month went by without a question.

Why doesn't a one-off purchase fit either?

A one-off purchase charges everyone the same price up front, while a twin's value sits in the one question each buyer brings. An ebook or a recorded course can work that way because you know what the buyer needs. With a twin, you can't know the question in advance.

One person needs two minutes. Another talks for twenty. A flat price charges the first one too much and the second one too little. With a pay-per-minute AI twin, each person pays for the minutes they actually use.

How do the three ways of charging compare?

Pay per minute is the only one of the three where each person pays for the time they actually use.

What do you keep when you charge per minute for your expertise?

You keep 80% of whatever you charge above the €0.50 base, so at €2.00 a minute you keep €1.20 of every paid minute.

Here's how that works for a consultant who bills €300 an hour for their own time. That's €5.00 a minute, and every one of those minutes sits on their calendar.

They build a twin from their written guides and the answers they keep giving new clients. They set its price at €2.00 a minute, well under their own rate. Here's what happens to each paid minute, with the platform base at €0.50:

• The visitor pays €2.00.
• The markup is €2.00 minus the €0.50 base, so €1.50.
• Our commission is 20% of the markup, so €0.30.
• The consultant keeps €1.20.
• The platform keeps €0.80, which is the base plus the commission.

More arithmetic from the same inputs:

• Once any free minutes are used up, a ten-minute conversation with the live avatar costs the visitor about €20, and the consultant keeps about €12.
• A single text question costs a quarter of a minute, so €0.50. The consultant keeps €0.30.
• Sixty paid minutes cost visitors €120 in total, and the consultant keeps €72.
• It takes 84 paid minutes to reach the €100 payout threshold.
• At €3.00 a minute the consultant would keep €2.00 a minute instead.

These sums only show how the split works on a price. How many minutes people pay for depends on your audience, and we don't have an honest figure for that. The price is yours, and you can change it from your dashboard whenever you like.

To try your own numbers, the calculator on our page about how to monetize an AI avatar works it out for any price. Consultants who want the whole picture can read AI twins for consultants.

Can a twin handle more than one conversation at a time?

Yes, your twin can handle hundreds of conversations at once, and each one has its own meter.

That's where it differs most from the consultant. Their €300 hour is one person at a time, and sixty paid twin minutes leave them €72, which is less than an hour of their own time. While the consultant is with one client, though, the twin can be in many other conversations, and none of those minutes come out of the consultant's week.

So what a twin earns is limited by how many people want to ask you something. Your calendar doesn't come into it.

It also changes which questions reach you. The short, repeated ones can go to the twin at any hour, and the ones that need a client's whole situation on the table still come to you at your own rate.

Where does a twin fit in the AI twin creator economy?

A twin sits beside what you already sell, such as a course, a cohort, a book or paid calls, and answers the questions that come up between them. Adding one doesn't mean dropping any of it.

• Someone reading your book gets stuck on chapter four at eleven at night.
• A cohort member wants a point from week two explained again before the next session.
• A prospect wants to hear how you think before paying for a call.
• A past client wants a reminder of a method you taught them last year.

Two things connect the twin back to what you sell. It answers from your material, so if your course or book is part of what you upload, the twin knows about it and can mention it when it's relevant. And you can add one link to a page of yours, such as a course you sell or your booking page, with a short line of text. A visitor who has run out of free time and credits sees it next to the option to buy more. The button has a fixed label, "Open the course" in English, and English-speaking visitors see your line of text above it.

The platform is for everyone: coaches, consultants, teachers, authors, doctors, psychologists and psychotherapists. Each professional decides for themselves, under their own professional rules, whether a twin fits their work. The twin shares what you've uploaded as general information. It doesn't give therapy, diagnosis or treatment, and it points people to you or another professional when they need one. Coaches can read how to build an AI coach for their side of it.

What does a free taster cost you?

At the default of 3 free minutes and 12 free questions, a visitor who uses the whole taster costs you about €1.50, or €1.56 if they spend it all on typed questions. You pay for it from your own credits at the base rate.

You set the free taster yourself, per twin: a number of free minutes and a number of free questions. When either runs out, the free part ends. Set either one to zero and there's no free part at all. New twins start with the default above.

Once the taster is used up, visitors buy credits to keep going. Your twin replies in each visitor's language.

How does the money reach you?

Your earnings are paid out through Stripe, monthly, once they pass €100. They build up with every paid second, and anything below the threshold carries over to the next month. We cover the Stripe payout fees.

Stripe checks your identity when you connect payouts. Your consent to cloning your face and voice is recorded and stored. No twin goes live without both.

You can switch your twin off any time. If you withdraw consent, we delete your cloned face and voice at our providers as well as in our database. More on pricing and the free taster: charging for your AI clone.

How do you build a pay per minute AI twin?

You upload your material, add a photo and a short voice clip, set your price and connect payouts. Building one is free, and visitors are only billed when they talk to it. For your material, you can upload documents (.docx), text files, audio or video, or paste text.

Create your twin, or read how it works for experts first.