Lispr now has a free plan and a Pro plan. Here is where we drew the line.
Lispr has been free for everyone since launch. It now runs on two plans, and the way we split them says more about the product than the prices do.
The standard move with a free tier is to cut features. English only. Three languages instead of thirty. The formatting held back as the thing you pay for. The free version gets built to be annoying enough that you upgrade.
We capped capacity instead. Every feature Lispr has sits on the free plan: push-to-talk dictation in about 99 languages with mid-sentence switching, instant translation into 32 languages, AI formatting, the personal dictionary, local history. No account needed. The limit is 60 dictation minutes a week and it resets every Monday. For most people that is several times what they actually speak into an app. None of it moves behind the paywall later.
Pro is $79 a year or $7.99 a month, with lower regional pricing in selected countries. It removes the minutes cap and carries across every device on one account. New installs get 30 days of Pro without a card.
The logic is simple. Someone who dictates ten minutes a day is not a customer. They are a user, and a user who never pays still tells other people the app works. Someone who dictates two hours a day hits 60 minutes on Tuesday morning and knows exactly what $79 buys.
There is also a referral link on your account page. Anyone you invite gets 30 days of Pro once they start actually using Lispr, and if you are on the free plan you earn Pro time back for it.
One question for anyone who dictates daily: is a weekly minutes cap the right shape for a free plan, or would you rather the limit sat somewhere else, like a per-session length or a monthly pool?
And if you have been using Lispr, an honest review on the product page is worth more than anything we can say about ourselves. Good or bad.
Replies
congrats on the launch. its a great product best of luck.
Lispr
Myroslav here, CTO at Codebridge and one of the makers. Adding the two questions that always follow a pricing change.
What happens when you hit the cap: dictation pauses until Monday. Nothing you have already dictated is touched or deleted, and your vocabulary list and history stay on your machine on both plans.
On audio: nothing is retained on our infrastructure. The transcription provider holds it briefly to check for abuse, up to 30 days under their policy. That is the accurate version and it is narrower than the blanket claim you see on a lot of dictation apps, so it is worth stating precisely.
Narrower question than Konstantin's, for anyone who would run into the limit: would you want to see the weekly cap coming, or is a hard stop you discover on the day easier to live with?
Been using this for a couple of weeks and the speed is genuinely surprising. What's running under the hood, and does it run locally or on your side?
Lispr
@khm Thanks, that's the part we spent the most time on. It runs on our side, not locally. Whisper large-v3-turbo, hosted, which is also why the app is small and there's no model file to download. Median from key release to text at your cursor is 346 ms, measured server-side on live traffic rather than in a demo.
Since server-side raises the obvious follow-up: your audio isn't retained on our infrastructure. The transcription provider holds it briefly to check for abuse, up to 30 days under their policy. Your vocabulary list and dictation history stay on your machine.