Free to deploy is a commitment, not a giveaway
Standing a deployment up is real work and real money: ingesting the archive, indexing it by meaning, building the branded surfaces, and paying for inference from the very first question anyone asks. We are glad to carry all of it for a business whose audience makes the threshold a realistic target, because that is a bet we expect to win, and because carrying it is the strongest signal we can send that we believe the thing works.
We cannot carry it for everyone who is curious. A deployment built for a business that decides three months later it was not for them is a straight loss, and the businesses who do commit end up paying for it. So where the audience is smaller, or where the intent is still exploratory, we ask for a one-time onboarding fee and a monthly minimum until subscriber revenue covers it. The amounts are modest and they are sized against your scale.
The point was never the money. The point is that both sides put something on the table before anyone spends six months on this. Where you fall is a conversation we have at the start, in plain terms, against your actual numbers.
The threshold is real
If your audience never reaches the subscriber floor, the platform earns nothing and we will eventually need to have a conversation about whether the deployment continues. We would rather set an honest threshold against your actual audience size than a decorative one.
It works on archives, not on ambitions
The engine needs a substantial body of published material to answer from. A handful of videos will not support cited answers or a personal guide. We will tell you if your archive is too thin, before any work starts.
Your audience has to be willing to pay
The whole model rests on end-user subscriptions. An audience that has never paid for anything is a genuine risk, and it is one we would rather size up with you honestly than discover six months in.