A catalog AI training course follows a standard program open to several companies; a custom course is built around your own use cases and delivered to your teams only. The right choice comes down to three factors: how many people you need to train, your AI maturity, and the level of proof you need, especially under Article 4 of the AI Act.
Read the full guide
AI Training for Business: The Complete 2026 Guide (Funding, Certification, EU AI Act)
Answer a few questions and get a personalized assessment with recommendations tailored to your industry.
Assess my AI maturityRelated articles
Dive deeper with these complementary articles.
AI training for businesses in France costs between €350 and €3,300 ex. VAT per person in 2026, with up to 100% OPCO funding for SMEs under 50 employees. Since August 2, 2026, the EU AI Act requires every company using ChatGPT, Claude, or Copilot to guarantee minimum AI literacy across its teams, with sanctions up to €7.5M or 1.5% of global annual turnover. Here is how to design a truly operational, funded, and compliant program.
A catalog AI training course follows a standard program open to several companies; a custom course is built around your own use cases and delivered to your teams only. The right choice comes down to three factors: how many people you need to train, your AI maturity, and the level of proof you need, especially under Article 4 of the AI Act. This comparison gives you a practical decision grid, not a list of providers. For the wider picture (funding, certification, compliance), see the complete guide to AI training for business.
Custom training is defined by content built for you, not by the fact that it happens on your premises. People often confuse "in-house" with "custom". An in-house session brings together only your staff, but it can run a catalog program word for word. True custom work starts upstream: an analysis of your use cases, the tools your teams already use, and exercises built on your own documents and processes.
A catalog course, by contrast, follows a fixed outline designed to suit as many people as possible. That is both its strength (proven content, dates available quickly) and its limit (generic examples, sometimes far from your business). Keeping this distinction in mind avoids the most common mistake: paying a custom price for standard content with a new label.
If you need to train one to three people or set a shared baseline, catalog is almost always the right call. Paying for a dedicated trainer for a single person makes no sense against an inter-company session already on the calendar. Short certified courses on the market usually run between 1,500 and 2,500 € per person, and intensive bootcamp formats between 1,500 and 7,500 € depending on length.
Catalog also shines on flexibility: you enroll whoever you want, on a date that suits you, without pulling a whole team off work the same day. It fits a first round of awareness, an executive who wants to understand the stakes before deciding (see where executives should start), or building a single skill such as writing effective prompts.
As soon as the goal shifts from understanding AI to transforming an actual workflow, custom takes the lead. Training ten sales reps to draft proposals with AI, twenty case handlers to process their email and files, a support team to answer faster: these goals require exercises on your real tools, your real documents, your real constraints. A fictional case study does not build that reflex.
The starting logic changes too. You are not training an individual, you are equipping an organization. So the right entry point is not the catalog but an audit of high-impact use cases, before the program is even set. That upstream work is why custom training is quoted rather than priced off the shelf: the content does not exist yet, it gets built. For ranges and funding levers, see our article on the cost of AI training for business.
| Criterion | Catalog training | Custom training |
|---|---|---|
| Content | Standard, identical for all | Built on your use cases |
| Indicative price | 1,500 to 2,500 € per person (short), 1,500 to 7,500 € (bootcamp) | On quote, based on audit and length |
| Time to start | Short, dates already scheduled | Longer, scoping phase upfront |
| Relevant group size | 1 to 3 people | From 6 to 8 people |
| Business grounding | Low to medium | High, on your real tools |
| Proof of competence per role (AI Act) | Hard to show | Easy to document |
| OPCO funding | Yes, if Qualiopi-certified | Yes, custom in-house included if Qualiopi |
This table does not pit a good format against a bad one. It shows that each option answers a different need. Many companies combine the two: a catalog module for general awareness, then a custom track for the roles where AI truly changes the output.
Price per person is misleading: the cost that matters is a skill actually applied on the job. A catalog course at 1,800 € that changes nothing in practice really costs 1,800 € plus the salary hours spent for no return. A pricier custom track followed by real adoption can be cheaper per result.
Three cost items often stay invisible at decision time. First, staff time during and after the training. Then travel and logistics for inter-company formats, on top of the tuition. Finally, and this weighs the most, the application rate: training with no business use case rarely leaves a mark on daily work. That is exactly what custom aims to secure, by working at the right pace and duration (see how many hours it takes to change practices).
Article 4 of the AI Act calls for a level of AI literacy suited to the context of use, which favors content tied to roles. Regulation EU 2024/1689, whose Article 4 has applied since 2 February 2025, requires organizations using AI systems to ensure a sufficient level of understanding among the people involved. The text says you must account for technical knowledge, experience, education, and the context of use. Supervision and penalties fall to national market surveillance authorities from 3 August 2026.
The regulation imposes no official program, no minimum hours, no certification. That flexibility is an advantage, provided you can document what each profile learned against their real usage. Custom training, tied to concrete cases by role, produces that proof more naturally than a generic module everyone takes the same way. For the detail of the obligation, see our analysis of Article 4 of the AI Act applied to training.
Three questions settle most cases. How many people are you training, and in which roles? Are you trying to understand AI or to change an existing output? Do you need to prove, role by role, a level of literacy under the AI Act?
If you train few people and aim for awareness, go catalog. If you train a whole team around a specific business use, or if you must document competence per profile, custom is justified. When in doubt, a short scoping check confirms the need before you commit a budget. If you go custom, write a brief: goals per role, tools to integrate, priority use cases, expected deliverables, and adoption metrics at six months. That document is the difference between real custom work and a catalog in disguise. When comparing providers, our article on the criteria for choosing an AI training provider gives you the grid to apply.
Is custom training always more expensive? At the sticker price, yes, because it includes scoping work. By result, not necessarily: with a better application rate, the cost per skill actually used can be lower than a poorly fitted catalog course.
From how many people does custom become worthwhile? Usually from six to eight participants sharing the same business use. Below that, an inter-company session costs less for an equivalent result.
Can custom training be funded by an OPCO? Yes. A custom in-house course delivered by a Qualiopi-certified provider is eligible for OPCO funding, just like a catalog course. The detail of the schemes is in our guide on OPCO funding for AI training.
Can you mix catalog and custom? Yes, and it is often the most effective approach. A catalog module sets the shared baseline, a custom track anchors usage in the roles where AI changes the output.
Is catalog enough to comply with the AI Act? It can support general awareness, but Article 4 asks for a level suited to the context of use. Proof per role is simpler to produce with content tied to the work.
Is a ChatGPT course catalog or custom? Both exist. A generic program on the tool is catalog; a track built on your documents and processes is custom (see our ChatGPT training for business).
GrowthPerf is a Qualiopi-certified training provider specialized in AI and automation for SMEs and nonprofits. We rarely start with a program. We start with a short audit of your use cases, used to say honestly whether a catalog format is enough or a custom track is justified. The goal is not to sell you the priciest format, but the one that leaves a mark on real work and documents cleanly against the AI Act.
If you are still unsure, the simplest step is a thirty-minute conversation. Book a free AI training audit: you leave with a clear recommendation, catalog or custom, and a first brief if custom is the way. To place this choice within an overall strategy, the complete guide to AI training for business remains your starting point.