When you ask a provider "how much does an AI training for our team cost?", you almost always get the same answer: it depends. That's annoying, but it's also fair — a one-hour webinar for 50 people and a program built on your company's processes for a team of 15 are different things, even if both are called "AI training." Knowing what moves the price is what lets you compare offers that look alike and aren't.

What actually moves the price

The first factor is the number of participants and the format. Online, for a large group, the per-head cost drops, and so does individual attention. On-site, for a small group working on their own cases, you pay more per person, but people leave with something they use the next day.

The second, and the one that matters most, is the degree of customization. A generic training — the same slides for every company — is cheap because the provider delivers it dozens of times unchanged. A program built on your workflows, tools and roles takes real preparation before the first session, and the price carries that work.

Then depth matters. A single session with nothing after it is a small expense and a matching impact. A program with exercises on real work, with deliverables and a few weeks of support afterward, costs more, because adoption needs that follow-up.

Finally, if you need EU AI Act compliance documentation or named certificates, those add to the cost, but they settle an obligation you'd have to tick off anyway.

The order of magnitude, without invented numbers

I can't give you a fixed price here, because it would be a lie — nobody serious quotes without knowing how many people are involved and what you want to achieve. But you can think in three tiers: an informational webinar is the cheapest and usually stays at the "we've heard about AI" level. A hands-on one-day training with exercises sits in the middle. A customized program over several sessions, with deliverables and support, is at the top — and it's the only one that genuinely changes how the team works.

As a rule of thumb, the closer it is to your real work, the more it costs and the more it shows.

The cost you don't put in the quote

The most expensive option is the cheap training nobody applies. You pay little, people sit for a few hours, go back to their desks, and work exactly as before. The money is small, but the team's time is lost, and the problem you paid to fix stays where it was.

The real cost of any training is the time of the people pulled off work. If you're pulling them off anyway, it makes sense for the sessions to produce something that stays — a library of prompts, a configured assistant, a plan — rather than just a day of presentations.

What to ask so you compare properly

Before you look at the price, put these questions to every provider: is it built on our processes or the same material for everyone? What do we walk away with, concretely? Is there support afterward, or are we on our own from month two? The answers explain the price difference better than any comparison table.

In the TVL Academy program we start from the company's real processes, and the team leaves with prompts, AI assistants configured per role, and an automation plan. We calculate the price based on the number of participants and the format — tell us in 30 minutes what you need and you get a clear quote, with no hidden costs.