Jumpspeak vs Duolingo: I checked the prices everyone quotes

The three pages ranking for "jumpspeak vs duolingo" agree on the central number. Jumpspeak costs $69.99 a year. One of them says $299 for lifetime access. Another says 17 languages, a third says 5.

On 21 September 2026 I opened jumpspeak.com/pricing. It offers $69 for 3 months and $79 for 12 months. Then I pulled the App Store listing for the same app, where Apple prints every in-app purchase Jumpspeak sells: Premium at $49.99, $59.99 and $79.99, Premium AI at $49.99, $89.99, $99.00 and $179.00, a Family Plan at $149.99, and Lifetime Access at $499.00.

Ten price points across two first-party sources. $69.99 is not one of them. Neither is $299.

The decision, in one paragraph

You're not choosing between two apps. You're deciding whether to stop using a free one and start paying, and the honest framing of that is $0 against roughly $79, because Duolingo's free tier is a real course that never expires and Jumpspeak's full library is behind a subscription.

So the default is Duolingo free, and you should pay Jumpspeak only for a reason you can name out loud. There is one good one: Jumpspeak makes you speak from the first lesson, and Duolingo mostly makes you tap. If you have already done a year of tapping and you freeze when a person looks at you, that's a real gap and a speaking app is a reasonable purchase.

Two things to check before you pay. Whether your language is one of the seven with an actual course. And where you buy it, because that decides whether the 100-day guarantee exists for you at all. Both are below.

Who I am and what I haven't done

I build Verbamor, an iOS app for language learners. So I sell something in this category, and you should weigh everything here against that.

I have not used Jumpspeak. I have not run its speech recognizer, taken a lesson, or subscribed. Every claim below comes from Jumpspeak's own live pages, its help centre and its App Store listing, all read on 21 September 2026, and I quote them so you can check each one yourself.

That's worth saying because the top-ranked page for this query claims eight months of hands-on testing and a decade of coaching, carries a February 2025 byline under a 2026 title, and renders its entire body in JavaScript so the served HTML contains almost nothing. Its price rows come from a comparison widget whose fields include "partner active" and "dummy partner." A verdict is worth what its sourcing is worth.

What the prices actually are

Here is everything I could read on 21 September 2026, source by source.

jumpspeak.com/pricing. Two plans: $69 for 3 months, and $79 for 12 months badged "LIMITED TIME OFFER." When I moved to close the tab, a popup offered $59.00 against a struck-through $179.88, labelled "Get 67% off - today only," over a ten-minute countdown.

The App Store listing (published by Rype, Inc.) prints the ten in-app purchases above, with no billing term printed against any of them. So $79.99 might be a year or might be something else; Apple doesn't say, and neither does Jumpspeak.

What that spread means for you: Jumpspeak's price is a moving target set by which door you walk through and whether you look like you're leaving. Any page quoting one figure as the price, mine included, is quoting a photograph of a countdown timer.

Duolingo is priced the same way its App Store listing has been for a while, and the listing is the only first-party source for the numbers. Super Duolingo appears at $9.99, $12.99, $83.99, $95.99 and $119.99, again with no term attached to any price point. The listing does state the trial in words: "try Super Duolingo for 14 days free."

And Duolingo Max, the tier that actually competes with Jumpspeak on AI conversation, is where I have to stop. duolingo.com/max publishes no price. The $30 a month and $168 a year you'll see quoted are reported figures I can't trace to a Duolingo page, so I'm not printing them as fact. One of the three ranking pages omits Max entirely while promising to judge "better results," which is the same gap handled worse.

The guarantee has a trapdoor

This is the paragraph worth the page. Jumpspeak's 100-day money-back guarantee is real, and it does not follow you into the App Store.

From Jumpspeak's own pricing FAQ, verbatim: "This guarantee can only be made if you purchased via our website. If you purchased on the App Store, they will handle payments and refunds."

So a buyer who finds Jumpspeak through an ad, downloads the app, and subscribes inside it has bought the product without the guarantee attached. You then have Apple's ordinary refund process, which is a different thing with different odds.

The help centre adds windows the marketing never mentions. Premium gets 100 days. Premium AI gets 30. Family Plan and live tutoring get 30. And a subscription renewal gets 24 hours from the renewal charge, which is the window that matters in year two, when you've forgotten you subscribed. After that, in the help centre's words, "all sales are final," with "no partial or prorated refunds."

Of the three pages ranking today, one buries this in its thirteenth section and one never mentions it. If the guarantee is the reason you feel safe buying, buy on the website.

Twenty languages, seven courses

Jumpspeak's own FAQ names 20 languages. The very next sentence takes most of them back: "We offer only AI Tutoring (not our full course) for Korean, Japanese, Polish, Swedish, Danish, Hungarian, Turkish, Norwegian, Vietnamese, Greek, Bulgarian, Ukrainian and Russian."

Subtract those 13 and you get 7 languages with a full course: Spanish, French, German, Italian, English, Portuguese and Dutch.

Which makes the SERP's disagreement funnier. Two pages say 17, one says 5, and the right answer depends entirely on what you count, which is a question all three could have answered by reading two consecutive sentences on the pricing page. If you want Japanese or Korean, you'd get AI tutoring and no course, and the page telling you "17 languages" wouldn't warn you.

Duolingo publishes learner counts per course on its own directory and teaches non-language subjects too, chess and maths and music among them. Its nine largest courses now run through CEFR B2, a Duolingo Score of 129, announced 22 April 2026. Any page still repeating "Duolingo only gets you to A2" is a year and a half out of date, and one ranking today implies exactly that.

One more difference nobody on this SERP prints. Jumpspeak's App Store listing says "Only for iPhone." No iPad, no Mac. Duolingo's listing covers iPhone, iPad and Apple Vision, and the course directory works in a browser. Jumpspeak's own FAQ agrees: "You will not be able to use the app on your computer."

The speaking claim, and the citation under it

Jumpspeak's pricing FAQ says its method is "3.5x more effective in learning retention than watching videos/listening to audio, according The National Training Laboratories."

That citation is the learning pyramid, and the pyramid has no paper under it. NTL's originating research was internal, was never published, and is described as lost, so the retention percentages rest on, in Wikipedia's summary, "an unknown methodology of unknown quality." A learner cannot check it because there is nothing to check.

Which is a shame, because the underlying instinct is defensible on real evidence. Kang, Gollan and Pashler tested it properly in 2013: learners who tried to produce a Hebrew word from memory before hearing it beat learners who simply repeated after a recording, on a spoken production test two days later, with no loss of pronunciation quality. Producing language beats echoing it. That result is narrow, picture-naming of single words rather than talking at conversational speed, and it is still a real finding with a DOI.

Duolingo's speaking exercises have their own limit worth knowing, and it cuts the other way. Its repeat-after-the-character and translate-out-loud items put the target sentence on screen before you speak, so the retrieval step is done for you. You're reading aloud. That's a genuinely weaker exercise than producing a sentence from a bare meaning prompt, and it's the strongest honest case for paying for a speaking app.

What I can't tell you is whether Jumpspeak's recognizer is any good, because I haven't run it. The top-ranked page reports that it accepted mispronounced words and rejected clear ones, which would be the failure that matters most: a recognizer that says yes when you're wrong teaches a beginner to say it wrong. I'd want that measured rather than anecdotal before I'd repeat it as fact, so I'm flagging it as the thing to test in your own first week rather than passing it on as a finding.

Side by side

Which one you should pick

Pick Duolingo free if you're starting, if you're price-sensitive, if you want a course that reaches B2, if you study at a desk, or if your language isn't one of Jumpspeak's seven. That covers most people reading this, and it costs nothing, which is why I'd rather say it plainly than bury it under a table.

Pick Jumpspeak if you can already read and follow your language but your mouth locks up, your language has a real course, and you buy on the website so the guarantee is yours. Give it the 100 days and actually use the refund if it doesn't move you.

Pick neither, honestly, if what you want is conversation with a person. An hour a week with a tutor on italki or Preply does something no simulated conversation does, and we wrote up what that costs in the cost of each method. An app is what you do between those hours.

Where Verbamor fits, and where it doesn't

Verbamor doesn't compete with either of these for the reader this page is mostly for. It teaches you nothing new. It records a lesson you already had with a tutor, pulls out the words and phrases you got wrong, and builds scheduled cards from them, in Spanish, French, Italian, Portuguese and Russian. It's iOS only, there's no free tier, and it's $19.99 a month or $129.99 a year with the 7-day trial on the annual plan. Both are US prices, read off the app's own plan definitions on 21 September 2026; the App Store sets a different number in other countries.

So if you're at the "should I stop using the free app" stage, Verbamor is the wrong purchase and Duolingo is the right one. It becomes relevant later, once you're paying a human being and losing most of each lesson by Friday. Different problem, further down the road.

Questions people actually ask

---

Sources

Every load-bearing claim Verbamor makes is traced to its paper on the research page.