Comparisons
Is Duolingo enough to learn a language?
Short answer: not on its own, and the reason is specific rather than general. It is better at its job than its critics admit, and that job is narrower than its users expect.
Worth knowing before you read on: we make a vocabulary widget, so we are a competitor of sorts. So is nearly everyone else answering this question — a striking share of the "honest Duolingo review" pages you will find are published by language apps hoping you switch to theirs.
Which is a reason to lean on what research has found rather than on anyone's opinion, ours included.
What the studies find
The findings are unusually consistent for this field. Work examining Duolingo's outcomes has repeatedly found real gains in vocabulary, reading, writing and listening, alongside limited development of oral communication — a pattern reported by Wang, Yang and Li in 2018 and corroborated by a systematic review from Lekkas, Cilingir and Papadopoulou in 2021.
That ceiling is not a scandal. Most courses are explicitly pitched at beginner to lower-intermediate, and getting a beginner to A2 reliably is genuinely difficult. The mismatch is between that and what daily use implies — a streak counter measures attendance, not progress.
Where it is strong and where it stops
| Skill | Verdict | Why |
|---|---|---|
| Vocabulary | Strong | Repeated exposure across many short sessions, which is what vocabulary needs. |
| Reading | Strong | Steady practice at sentence level, well matched to the exercise format. |
| Listening | Good | Regular audio, though usually slow, clear and studio-recorded rather than natural. |
| Writing | Moderate | Mostly translation of given sentences rather than composing your own. |
| Grammar understanding | Weak | Patterns are absorbed by exposure; explanations of why an answer was wrong are thin. |
| Speaking | Weak | The consistent finding across studies. Little unscripted production, and no interlocutor. |
Duolingo changes its exercise formats regularly, so treat this as the shape of the tool rather than a permanent specification.
Recognition is not production
The common criticism — that Duolingo teaches you to translate rather than to speak — is really a point about exercise design. Tapping word tiles, choosing between options and translating a sentence you have been given all ask you to recognise or rearrange language. Very little asks you to generate a sentence nobody has shown you.
That produces the experience many long-term users describe: understanding far more than they can say. It is the same passive-versus-active gap that shows up in vocabulary generally, which we went into in how many words you need to speak a language.
The related weakness is explanation. When an answer is wrong, you learn that it was wrong more reliably than you learn why, so grammar tends to be inferred rather than understood. Some learners are entirely happy with that; those who want rules will want a course or a teacher.
What it genuinely does better than the alternatives
Duolingo solves the problem most language learners actually fail at, which is not method selection but showing up. Streaks, leagues and short lessons are easy to sneer at and they work: a year of daily practice from an imperfect tool beats three weeks of an excellent one.
It is also free at a level that is genuinely usable, and it covers languages that paid competitors do not bother with. For a complete beginner with no idea whether they will stick with a language, it is a sensible first move and a cheap one.
What to add, and when
The useful question is not whether to abandon it but what to put alongside it. Add whichever of these matches what you are missing.
- Speaking with a person. A tutor, a conversation exchange, anything with an interlocutor. This is the gap the research identifies most consistently, and no app fully substitutes for it.
- Real input. Graded readers, then native podcasts and television. Duolingo's audio is deliberately clear; actual speech is fast, accented and full of things textbooks omit.
- Grammar reference. A single good grammar book answers in a page what a hundred wrong answers will not.
- Retention between sessions. Words met once and not seen again fade within weeks, whatever app introduced them — the subject of why you forget vocabulary.
That last one is where Glancealot fits, and it is worth being precise about how small the claim is: it puts words on your lock screen so they come back without a study session. It does not teach grammar, it cannot test your recall, and it will not get you speaking. It addresses the retention gap and nothing else.
If you would rather compare full courses than supplement one, we looked at the main options in the best language learning apps.
The verdict
Duolingo is enough to build a habit, a few thousand words and a working grasp of how a language fits together. It is not enough to make you conversational, because conversation requires producing language under time pressure with another person, and no amount of tapping tiles rehearses that.
Used as the daily engine with one thing bolted on for speaking, it is a perfectly good centre of a learning routine. Used alone and indefinitely, it produces a very well-informed reader who freezes when someone talks to them.
Frequently asked questions
On its own, generally not. Studies consistently find gains in vocabulary, reading and listening but limited development of speaking, and most learners plateau somewhere around CEFR A2 to B1 — basic conversational ability rather than fluency. It is a strong beginner tool and a poor complete course.
Most courses are pitched at beginner to lower-intermediate, and research comparing outcomes against the CEFR suggests learners typically plateau at A2 to B1. Reaching B2, the level most people mean by fluent, generally requires speaking practice and material Duolingo does not provide.
Two things, both genuinely valuable: building a daily habit, which is the thing most learners fail at, and drilling vocabulary through short repeated sessions. Its gamification is often mocked but it solves the consistency problem better than most alternatives.
Because most exercise types — tap-the-pairs, multiple choice, translating a given sentence — ask you to recognise or rearrange language rather than generate it. That builds recognition well and production poorly, which is why learners often understand far more than they can say.
Add whatever supplies what it lacks: a tutor or conversation exchange for speaking, graded readers or native media for natural input, and a spaced repetition tool for retaining vocabulary between sessions. Adding one of these matters more than switching apps.
Clearly yes, and that is worth saying plainly. A learner doing Duolingo daily for a year will know substantially more than one who intended to take a class and never did. The criticism is about its ceiling, not its value.
