What credits buy: one currency for the whole platform

Usage-based pricing for AI products usually means one thing: a bill that is a mystery until it arrives. You ran some articles, made some keyword queries, and at the end of the month a number appears that you cannot reconstruct or predict. We did not want to build that, so Maark bills in credits — one currency for everything the platform does, with a published price for every action.

This post is about the model, not the numbers. The prices themselves live on the pricing page, which is the source of truth; we are not going to quote figures here that could drift out of date.

One currency, one published rate card

Every billable action has a fixed, published price in credits: producing a full article, producing a quicker one, seeding a Keyword Universe, running an AI-visibility scan, refreshing a crawl, checking rankings, assessing backlinks. That price is on a rate card you can read before you act, so you always know what something costs before you run it — no surprise line items.

The point of a single currency is flexibility. Credits are not earmarked per feature: budget you do not spend on content can go toward keyword research, or an AI-visibility check, or a site audit. You are buying platform capacity, not a fixed basket of one thing.

And because the price is the card's price, an article you did not want costs exactly what the card says — not whatever the underlying models happened to charge that day. The rate card is the contract; the volatility of model costs stays on our side of it.

Grants versus top-ups

There are two ways credits arrive, and they behave differently on purpose:

• A monthly grant. A paid plan grants credits when you subscribe and again at each renewal. These are your included monthly allowance. Granted credits do not roll over — each cycle starts fresh with a new grant.
• A top-up. Credits you buy as a pack on top of your plan. These roll over, staying spendable for up to a year, because you paid for them directly and they should not evaporate at a cycle boundary.

When you spend, Maark draws down your expiring monthly grant first and only then touches your longer-lived top-up balance — so the credits with a deadline get used before the ones without one. It is the order you would choose yourself if you were watching the meter.

When the balance runs low

If your balance approaches empty, Maark warns you — an in-app banner and an email around the point where you are running down your monthly grant — rather than letting you discover it mid-task.

If it reaches zero, the product pauses gracefully instead of failing. A billable action shows a top-up prompt instead of a bare error, and everything resumes the moment you top up or your cycle resets. Nothing is lost in the meantime. For teams that would rather not think about it, there is an optional auto-refill: it tops up when you cross a threshold, within a monthly cap you set, and it only ever runs with a card you have explicitly put on file for it. Hit your cap and it simply stops — behaving like a zero balance, never a runaway charge.

Why credits instead of per-seat or per-article pricing

A per-seat price punishes you for adding teammates who barely log in. A rigid per-article price forces every job into the same box whether it is a flagship guide or a quick update. Metering real work against one currency lets the price track the actual cost of the actual thing you did — an expensive action costs more credits, a cheap one costs fewer — while keeping the whole system legible: one balance, one rate card, one place to look.

Agencies get the same model at portfolio scale: one pooled balance across every site you manage, so a busy month on one client can draw from the same well as a quiet month on another.

For the exact prices, credit amounts, and plan allowances, see the pricing page — and the billing help articles go deeper on how grants, top-ups, and auto-refill behave day to day.