How we price: the economics of a humanizer and what you're paying for

Pricing pages usually hide the reasoning. Here's ours, because we'd rather you understood the deal than felt tricked by it.

What actually costs money

Two things scale with your usage: the rewriting itself (a call to the engine behind each humanize, priced roughly by length) and detector scans (each check has a cost). Everything else (hosting, auth, payments) is comparatively fixed. Because the variable cost tracks the number of words you process, we price by words. That's the honest unit; charging by 'requests' or selling 'unlimited' either penalises short edits or forces us to police heavy users with hidden throttles.

It's also why unused words don't roll over: the cost is incurred when you use them, not when you buy the plan, and rollover would just be a more confusing way to price the same thing.

Why failed requests are free

If the engine times out or errors, we didn't deliver the rewrite, so you shouldn't pay for it. We build that into the meter (words are reserved and released on failure) rather than making you ask for a refund. It costs us the occasional wasted call; that's the right side to err on.

The ladder

Free (300 words a month) is there so you can actually judge the tool on your own writing, not a demo paragraph. Starter, Pro, and Max step up the monthly allowance and the per-request cap, and add the multi-detector view and automatic splitting of long text. The 24-hour pass exists for a deadline; credits exist for occasional use without a subscription. In Ghana, all of it is payable with mobile money through Paystack, because card-only pricing excludes most of the people who use us there.

We set Starter slightly above the cheapest competitor on a like-for-like basis and undercut on the middle tier, and we're comfortable saying so. You're paying for a careful edit loop with a detector in it and billing you can predict, not for a promise no tool can keep.