Documentation
Understand how Skilzy works, in a few minutes.
Skilzy gives you a monthly AI quota to chat with the AI and generate images or videos. This quota refills automatically every month, on the anniversary date of your sign-up. Paid accounts and licenses issued by a training organization get a higher quota. When your monthly quota runs out, you can buy Skilzy credits: universal credits that take over so you can keep generating images and videos.
By completing lessons, you earn XP (experience points) that level you up. Gems are a currency you accumulate as you progress and spend in the Shop. Your streak rewards consistency: come back every day to grow it. And every week, your weekly goal sets the pace: hit it and you finish your programme in 30 days.
A lesson is a series of short steps: explanations, interactive exercises and quizzes. The last lesson of each module is a "boss": it validates what you learned in the module. At the end, your performance is summed up with a star score — the more exercises you get right on the first try, the more stars and XP you earn.
Some programs prepare you for an official certification, registered with France Compétences' Répertoire spécifique ("RS…" codes) and eligible for CPF funding. You follow the modules at your own pace, then take a mock exam in the format of the final exam. A training certificate is generated at the end of your path.
Tutoring means live class sessions, by video call, led by a trainer. If your plan includes Tutoring, the "Tutorat" tab shows the session calendar: you book your spot, then join the call on the day. Tutoring is offered through partner training organizations.
You can invite other people to join Skilzy with your personal referral link. When a referee signs up and moves to a paid plan, you get a reward. Find your link and track your referrals in your settings, under "Referrals".
Prompt — the instruction you write to an AI to get a response. Model (or LLM) — the AI's "brain", trained to understand and generate text (Claude, GPT…). Token — the unit the AI uses to split text; a message's length is counted in tokens. Hallucination — when an AI produces an answer that is wrong but stated confidently. That's why you should always double-check. Vibe coding — building apps by describing what you want to an AI, instead of hand-coding everything.
