Babbel vs Duolingo: ask about the ceiling

I have not used either app, and this page is built entirely from what Babbel and Duolingo publish themselves. I fetched every page cited here on 21 September 2026 and I quote it with that date. I also sell a language app, which you should factor in while reading. The upside of doing it this way is that you can click every claim and check it, which is more than I can say for most of what ranks for this search.

The answer, in one paragraph

Buy Babbel if you want grammar explained to you in sentences. Buy Duolingo if the thing that has stopped you before is not showing up daily, because its free tier is real and its habit machinery is the best in the category.

Now the part that should actually decide it. On 22 April 2026 Duolingo announced that nine courses teach through CEFR B2, and described what changed in its own words: "Previously, most of these learners could only complete content up to A2 (Duolingo Score 59)." The nine are English, Spanish, French, German, Italian, Portuguese, Japanese, Korean and Chinese. Duolingo advertises 40+ languages, so most of its courses are outside that list.

Check your language against those nine before you compare anything else. It is a bigger difference than any feature on either app.

The ceiling each company publishes

Duolingo is unusually specific here, and it deserves credit for that. It maps CEFR to its own Duolingo Score, puts B2 at 100 to 129, and describes what a 129 buys you: you are an "independent user" who can "express your viewpoints on a wide range of topics," while you "may still have to search for your words occasionally."

That is an honest description of B2, including the hedge. It is also a claim about what the course contains rather than what learners come out with, and those are different things.

Babbel publishes a ceiling too, in a quieter place. Its Spanish course page lists its own levels against CEFR: Newcomer A1, Beginner A2, Intermediate B1, Upper Intermediate B2, Advanced C1. That is a rung above Duolingo's stated B2, on paper.

I went looking for the same ladder on Babbel's French, Italian, German, Russian and Portuguese pages and did not find it, so treat A1 to C1 as a published Spanish claim rather than a company-wide one. One subscription covers all 14 languages, and Babbel's about page counts "60,000 lessons across 14 languages."

Two caveats travel with all of this. Neither number is a measured learner outcome, and both are the company grading its own homework. The strongest published result on app learning is still Jiang, Peters, Plonsky and Pajak (2024), which followed 245 learners through the Basic content of a Duolingo English course and found they scored Intermediate High in reading and listening. Three of its four authors work at Duolingo, and the study measured reading and listening only. Nobody has published the speaking equivalent.

What they cost, and why I won't give you a table of totals

This is where the pages ranking for this search quietly fall apart, so let me show you exactly what is knowable.

Duolingo. The US App Store listing for Duolingo on 21 September 2026 showed Super Duolingo at five price points: $9.99, $12.99, $83.99, $95.99 and $119.99. Apple attaches no billing term to any of them. There is no honest way to tell you which is the monthly rate and which is the year, so I won't. The same listing offers "Super Duolingo for 14 days free."

Note what that means for the competition. A page that prints "Duolingo Super, $6.99/month" is quoting a figure that is not on Duolingo's own store listing today. One of the three pages ranking for this query still does.

Babbel. Babbel's own pricing page builds its plan table in your browser, so fetching it returns an empty shell, which is why so many pages carry stale Babbel figures. What I could read first-party was Babbel's App Store listing, on the same day, and it shows something the other comparison pages miss entirely: Babbel sells per-language plans and an all-languages plan at different prices.

The sample Apple displayed included a single language at $17.99 for one month, $45.99 for three months, Mexican Spanish at $89.99 for a year, and an All Languages 12-month plan at $107.99. Apple shows a sample rather than a full catalogue, so treat those as real observed points rather than the complete menu.

So no, you will not find a tidy twelve-month total for both apps on this page. Getting one would mean inventing a billing term for Duolingo and a plan table for Babbel, and a made-up number that looks precise is worse than an honest gap. Open both listings in your own store, where the term is attached to the price and the currency is yours.

One thing worth knowing before you click: Babbel runs a discount banner close to permanently, so an introductory rate and a renewal rate are two different numbers. Whatever you see on the way in is not what you pay in month 13.

Free is a price, paid in interruptions

Duolingo's free tier is real, gives you the full core course, and is one of the more generous things in consumer software. Any comparison that waves it away is not being fair.

But scoring it 10 out of 10 on value because it costs nothing, which one of the pages ranking today does, mistakes the meter for the bill. The free tier is metered in Energy, a battery that lessons consume and that you refill by answering correctly, watching an ad, or spending gems. Super's pitch on the App Store listing is "no ads" plus "Unlimited Energy" and "Monthly Streak Repair," which tells you what free actually costs: your attention, and your session ending when the app decides rather than when you do.

Duolingo does not publish the starting Energy amount. Third-party pages that quote you a specific number are quoting something Duolingo has never said.

Babbel's free offer is a different shape and it gets misdescribed constantly. The first lesson of every course is free, permanently, and Babbel states no trial length in days anywhere on its own pages. Any page telling you Babbel has a 30-day free trial is repeating an offer the company does not currently publish.

The thing neither one asks you to do

Here is the axis both apps genuinely share, and it is the one I care about most.

Duolingo describes its own speaking exercises plainly: they "prompt you to repeat words, translate sentences out loud, or take part in bite-sized dialogues," plus a microphone option on many writing exercises where you "speak your answer instead of typing it." Babbel runs speech recognition on your pronunciation, on a prompt you can see.

In both cases the sentence is on the screen before you say it. Your mouth does the work and your memory sits it out.

That distinction has a literature behind it. Kang, Gollan and Pashler (2013) taught learners Hebrew nouns two ways: one group heard each word and repeated it, the other tried to produce it from memory before hearing it. Two days later the retrieval group had better comprehension, better production, and no loss of pronunciation quality. Repeating after a model is the cheaper operation and it buys less.

Be careful how far you push that, though. The test was naming pictures of single words in isolation, which is not the same as finding a word mid-sentence while somebody waits. I am not going to pretend otherwise, because the study does not.

Babbel's review has a related limit it states itself: the vocabulary in your review is predefined by the lessons you complete, and you cannot manually add words to it. So the words you fumbled in a real conversation last week cannot get in. Its review ladder is fixed too, six steps and then it stops, with no algorithm named on any Babbel page.

Side by side

Every cell below traces to a company page I fetched on 21 September 2026.

Which one to buy

Work down this list and stop at the first line that describes you.

Your language is outside Duolingo's nine. Then Duolingo has not claimed B2 content for it, and if Babbel teaches your language at all, it is the better structural bet. This is the single most decision-relevant line on the page and it is the one no competing page prints.

You want the grammar said out loud. Babbel. It explains rules; Duolingo mostly has you infer them from examples, which works for some people and infuriates others.

You have never made it past week 2. Duolingo, free, today. Fix the showing-up problem with the tool built for it, and spend nothing while you do.

You want to talk to a person within a year. Neither, on its own. Both are teaching you to recognise and perform, and a conversation asks you to produce. An hour a week with a tutor from a marketplace does something no course app does, which is put you under mild social pressure with no sentence on the screen.

You already pay for lessons. Skip both. A course app teaches its own syllabus and cannot know what your tutor corrected on Tuesday, which is the wrong shape for the problem you have.

And the concession I actually mean: if you have lessons and you will type your own cards, use Anki. It is free on desktop and Android, its scheduler is excellent once you switch FSRS on, and it will hold your tutor's words better than either app on this page. The catch is that you build every card by hand, forever, and most people quietly stop around week 3.

That week-3 lapse is the only thing I built Verbamor for. It records your lesson, pulls the words you actually got wrong out of the transcript, and makes the cards so you don't have to. Five languages, iOS only, $19.99 a month or $129.99 a year with a 7-day trial on the annual plan, and no free tier. 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. If you don't take lessons it has nothing to work with, and one of the two apps above is the right purchase instead.

Questions people actually ask

---

Sources

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