Our entire pricing page is one number. Five dollars a month, or fifty a year. No seats, no tiers, no minutes bundle, no "contact us".
That's an unusual amount of restraint for software, and it wasn't restraint so much as arithmetic. Here's the reasoning.
What a captioning tool actually costs to run
Three things cost real money per video, and only three:
- Transcription. Priced per hour of audio by the provider. It's the largest variable cost and it is genuinely small.
- Rendering. Burning captions into the frame, which runs on Lambda and bills by the second.
- Storage and egress. Your source video, its thumbnail, and the export.
Everything else — the editor, the fonts, the presets, the transcript panel — costs the same whether one person uses it or ten thousand do. It's software we've already written.
That shape is why the plan is unlimited. A metered plan would mean building metering, explaining metering, and having an argument with someone about metering, in order to manage a cost that is a rounding error next to the support time the argument consumes.
Why not more
We could charge fifteen. Plenty of tools in this category do, and some of them are worth it.
The problem with fifteen is that it puts you in a bracket where people compare you to things with teams behind them. At five, the decision is small enough that the honest question is "is this useful?" rather than "is this better than the four alternatives I should now go and evaluate?" We would rather answer the first question.
There's a second effect, which is that a low price is a forcing function on support. At five dollars a month you cannot afford a support process that involves anyone doing anything twice. So the product has to be self-explanatory, the errors have to say what to do, and the export has to work the first time — because the alternative isn't expensive, it's impossible.
A cheap product can't afford to be confusing. That's a constraint we volunteered for and it has made the software better.
Why not less
Free would be worse, and not for the obvious reason.
Everything here that costs nothing to run is free — the whole editor, all the fonts and presets, unlimited videos, the transcript, the six browser tools. You pay when you export, because exporting is the part that spends money.
A free tier that included exports would mean rate limits, watermarks, or a queue. All three make the product worse for the people paying, in order to subsidise people who aren't. Gating the one expensive action is the least annoying line we could find.
Why yearly is fifty and not sixty
Five times twelve is sixty. Yearly is fifty — ten months' money for twelve months of the thing.
Two months is what paying up front is actually worth to us: it removes twelve chances to churn and twelve card transactions, and the discount is us sharing that saving rather than dangling a hook. It isn't larger than that because a half-price year would be a number chosen to look generous, and someone has to keep the renders running for the other six months. It's also a genuine commitment in the other direction — take a year's money and you owe someone a year of the thing continuing to exist.
What we won't build
No enterprise tier. No per-seat pricing. No AI avatars, no dubbing, no translation-as-a-service. Not out of purity — those are real products people want — but because each one turns a tool into a platform, and platforms need the pricing page we deliberately don't have.
If the thing you need is captions on a video, five dollars is the whole conversation.