How far the free route actually goes

Some links on this site earn a commission. This post is about the ones that do not, and it exists because a recommendation engine that never says "spend nothing" is not a recommendation engine.

What free actually covers

Spaced repetition: solved, free, and better than the paid version.

Anki is free on desktop and Android, open source, and its scheduling is better than anything bundled into a subscription app. It earns us nothing. It is in our recommendation set permanently, and the rule that put it there — that every result must include at least one zero-commission option — is enforced by a test rather than by good intentions.

The catch is real and worth stating: Anki decides nothing for you. You supply the words, the sentences and the discipline. A paid app's actual product is not its algorithm, it is the curriculum — someone else deciding what you learn next. That is worth paying for if you will not do it yourself, and it is the honest case for any app subscription.

Comprehensible input: free, and the bottleneck is finding it, not paying for it.

Language Reactor and similar tools make video watchable at a level below your own. Also free, also earning us nothing.

Grammar reference: free. For every major language there is a good free reference online. Paid courses package explanation with sequencing; the explanation itself is not scarce.

Where free stops

At roughly the same place self-study generally stops, and for the same reason.

What no free resource supplies is someone who will correct a sentence you have just invented. That is the entire content of the levels above B1, and it is the one thing that is genuinely scarce. Everything else in language learning has been commoditised; corrected output has not, because it requires a person's attention in real time.

Our model puts structured self-study at a 25% overhead at A1 and 70% at C1. Free resources sit at the same shape or slightly worse. That means:

• To A2: free is competitive with paid. The overhead is small and the price difference is total.
• To B1: free works, slower. Most of the gap is curriculum, which discipline substitutes for.
• To B2 and above: free alone is a long detour, and the thing missing is not information.

The recommendation, honestly

If you have not started, start free. Most people who intend to learn a language stop within two months. Buying a course before you have proven to yourself that you will show up is buying an outcome you have not tested. The free stack is good enough to find that out, and it costs nothing to abandon.

If you have a habit and a target above B1, buy the person, not the app. The scarce input is correction. A tutor at a few sessions a month costs less than several stacked subscriptions and supplies the thing subscriptions cannot.

If you know you will not build your own curriculum, a paid course is a fair trade. That is a real service and it is what you are paying for. Just be clear that it is what you are paying for.

Why a site funded by referrals publishes this

Because the alternative is worse for us.

Our matching rules were written before any merchant existed, and the ranking code cannot read commission rates — a test rewrites every payout to extreme values and asserts the output is unchanged. That is checkable and it is checked. But no test can make a reader trust the recommendations.

The only thing that can is the recommendations being right, including when right means "not yet". How this site is built and paid for sets out the rest, and the calculator carries no affiliate links at all.