Frequently asked questions

Whether it will work for you

I have used apps for years and still cannot speak. Will this actually be different?

It is a different activity, so it fails differently at worst. Tapping the right answer from four options is recognition, and speaking is production, and they transfer to each other weakly. Here every session is an unscripted spoken conversation from the first minute. Whether that works for you is mostly a question of whether you will do it, which is why there is a free trial rather than an argument.

I am a complete beginner. Is this too advanced?

No. Start as simply as you like and lean on English whenever you are stuck; the tutor meets you where you are and will not pretend you said something you did not. That said, if you have literally no vocabulary at all, a few weeks with something free first will make the conversations more useful, and we would rather say so than take your money early.

Is an AI tutor actually as good as a human teacher?

No, and anybody telling you otherwise is selling something. A good human teacher is better. What an AI tutor has is availability at eleven at night, infinite patience, no scheduling, no embarrassment, and a much lower price per hour. For a lot of people the real comparison is not against a human teacher, it is against the speaking practice they were not doing at all.

What if I freeze and cannot think of anything to say?

That happens and the tutor is built for it. It will offer you something to react to, hand you the line you were reaching for, or change the subject, in the way a person would. You can also pick a focus before you start, or take one of the suggested session cards, so you are never facing a blank conversation.

How it works

Does the tutor actually hear my voice, or does it read a transcript?

It hears the audio. Your voice goes to the model directly and it replies in audio, with nothing transcribing you into text along the way. That is unusual in this category and it is the main technical reason the app costs what it does. It matters because most of what marks a learner survives transcription intact: a Japanese vowel held too short, a missing aspiration in Hindi, a Thai tone on the wrong contour, the hesitation before a verb. A transcript hands the tutor the word you meant and it never finds out.

Can I interrupt it?

Yes, and you should. It stops when you start talking, the way a person does.

Does it correct my mistakes?

It corrects the way a good teacher does, which is mostly by feeding the right form back inside its own reply rather than by stopping the conversation to mark you. Being interrupted for every error is how learners end up afraid to speak. What you keep getting wrong is tracked and quietly shapes later sessions.

Is there a curriculum, or is it just chatting?

There is a written CEFR curriculum for every language, A1 through C1, of roughly 300 items each. What gets tracked is capabilities, meaning concrete things you can do like ordering a coffee or arguing a point of view, rather than grammar points. Each carries a strength that decays if you stop using it, so nothing is marked learned and filed away, and things you are losing get worked back into conversation when they fit.

What is Travel Mode?

Tell the app you are going somewhere the language is spoken and it changes shape for the trip. Before you leave, sessions rehearse what you are about to have to do. While you are there, you get one real-world mission a day, drawn from your own syllabus, pitched at or below your level and set in the city you are in: order something and pay for it, ask a shopkeeper where something is. You rehearse it, go and do it, then come back and tell your tutor what happened. Completed missions become stamps in a passport for that trip.

Do I need to configure anything?

No. Pick a language and talk. There is no placement test, no settings to tune, and no study plan to build. The app works out where you are from the conversations themselves.

Languages

Which languages can I learn?

Seventeen: Arabic (MSA), Dutch, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese (Brazil), Portuguese (Portugal), Russian, Spanish (Spain), Spanish (Latin America), Thai, Turkish and Vietnamese.

Why are Spanish and Portuguese listed twice?

Because they are treated as separate languages rather than as an accent setting. European and Latin American Spanish differ in the plural you, in whether c and z are a th sound, and in a lot of everyday vocabulary. European and Brazilian Portuguese differ even harder, because Portugal swallows its unstressed vowels and Brazil does not. Each variety has its own tutors, its own vocabulary and its own record of what you know.

Can I learn a language written in another script?

Yes. Japanese, Korean, Russian, Hindi, Thai and Arabic all have a Latin reading shown beneath the native script by default, so a beginner is never locked out of following along. It can be switched off in a moment once you can read without it, and it starts on because somebody who cannot read the script has no way to discover the setting exists.

Will you add more languages?

Slowly, and only where we can do them properly. Each new language needs a written CEFR curriculum, five tutors with genuinely distinct regional accents, pronunciation patterns authored for that language, and confirmation that the underlying speech models support it well. We would rather have seventeen that work than fifty that mostly do not.

Practicalities

What does Well Versed cost?

Standard is US$39.99 a month or US$399 a year. Premium is US$74.99 a month or US$749 a year and gives you double the conversation time. Both include every language and every feature. Every plan starts with a 7-day free trial and nothing is charged until it ends.

Why is it more expensive than other AI language apps?

Because a live audio model runs for the whole length of every conversation and is billed by the minute, which is a real per-minute cost that a text-based app does not have. The comparison we are built against is not a cheaper app, it is the hour a week you might otherwise spend with a private tutor at forty to sixty dollars an hour.

Is it unlimited?

No. Plans carry a cap, the cap is on spend rather than on a fixed number of minutes, and it runs on a rolling weekly window. The app shows you where you are and tells you plainly if you reach it. We deliberately do not print a minute figure next to the price, because any figure would be an estimate rather than a promise.

Which devices does Well Versed run on?

Well Versed runs on iPhone and iPad, with Android coming shortly. The App Store listing is the live answer if you are reading this later.

Do I need to talk to a real person at any point?

No. There are no human tutors, no language exchange partners and no group classes. It is you and an AI tutor, which for a lot of people is exactly what makes speaking feel safe enough to start.

How do I cancel?

In your iPhone settings, in about fifteen seconds. Subscriptions are handled by Apple, so we could not make it difficult even if we wanted to.

Privacy

What happens to my conversations?

They are stored against your account so the app can remember what you have practised, and they are never sold or shared. The AI providers we use do not train on them. You can delete your account and everything in it from inside the app, and that removes the lot.

Who else sees my voice data?

Three processors, and the app will not let you start until you have agreed to it: Google for the conversation model, ElevenLabs for transcription, and Daily for carrying the audio. They are named in the app before your first session and listed at wellversedlearning.com/privacy.

Is it suitable for children?

It is rated 13+. The conversation is open-ended rather than scripted, so it covers ordinary adult life the way any conversation course does, and the tutor writes its own replies rather than reading from a script. There is a full explanation on our age suitability page.