Volpiko
On réveille ton coach. Encore une seconde, et tu parles.
Comparisons · September 19, 2026 · 8 min read · by Jérémie Chiari
Yes, you can end up speaking Spanish with Rosetta Stone, but not from the daily lesson alone. The method is genuinely good at what it does: you look at a picture, you hear a phrase, you match one to the other, no translation in sight. TruAccent listens to you say it back and scores how close you land to a native speaker. Six months in, you recognize patterns fast, you read a menu without stopping, your accent has improved. Then a coworker asks you something you didn't see coming, and the silence comes back.
That's not a flaw in the tool. It's what a matching-and-repeating exercise produces, and that's not nothing. Speaking, on the other hand, is trained by speaking: out loud, every day, with someone who answers. The two don't replace each other, they add up. So the choice of a subscription comes down to one concrete question: what are you actually doing during the ten minutes you spend on it.
Rosetta Stone solved a problem most language apps still get wrong: it makes you trust the sound of your own voice.
The method is public and it hasn't changed much since the company started teaching languages this way in 1992: no translation, ever. You see a picture, you hear the target language, you connect the two, and your brain builds the association directly instead of routing through English first. It sounds slow at the start. It isn't decorative — it's the reason people who've tried a dozen apps say Rosetta Stone is the one that made a sentence pattern finally click without them memorizing a grammar rule.
TruAccent is the second piece, and it's underrated. Every phrase you repeat gets compared, in real time, to a bank of native and non-native speaker recordings, and you get a score you can act on immediately. That's real correction, even though it happens through your microphone instead of a red pen.
The habit engine works too, in its own way. Units are short, lessons unlock one at a time, and the app nudges you back with a streak. It isn't a daily email landing in your inbox, there's no story, no recurring characters, no jokes waiting for you on Tuesday morning. But the underlying idea is the same one other daily-lesson apps built their reputation on: keep the friction low enough that you actually open it again tomorrow.
Say it plainly: a method that gets you back every day for a year beats a brilliant method you quit after three weeks. On that front, Rosetta Stone earns its place. The problem is elsewhere.
An exercise that scores what you repeat trains recognition and pronunciation; speaking with someone, on the other hand, is a different skill entirely.
The Common European Framework of Reference, published by the Council of Europe, doesn't lump these together. It separates reception, production, and interaction. Cambridge English publishes the same breakdown, level by level. You can score well on reception and still freeze in interaction.
It's the classic story of someone who's kept a Duolingo streak alive for years, or spent a year matching pictures on Rosetta Stone. They can follow a slow conversation. They can't answer a question they didn't see coming.
Speaking fast isn't a knowledge problem. It's an access problem. In conversation, you get about a second to find a word, pick a tense, and commit to the sentence. A matching exercise gives you all the time in the world, and usually half the sentence already built. It's not the same mental move.
During ten minutes, depending on the format, you're not training the same thing:
None of the three replaces the other two. The mistake isn't choosing the first one. It's assuming it will produce the third one by itself. We go into this gap in more detail in I understand Spanish but I can't speak it.
One column matters if your problem is speaking: what you produce out loud, every day, without booking a slot.
The prices below are what the publishers themselves listed in 2026, before promotions. They move: several of these companies test their pricing constantly, and marketplaces bill by the tutor. Check the official page before you pay.
| Tool | Daily spoken conversation | Human or AI | Languages | Entry price (2026, verify) |
|---|---|---|---|---|
| Volpiko | Yes: the coach calls you, 10 min a day | 100% AI, no human | 6 open (English, Spanish, Japanese, German, Korean, Italian) | from €9.90/month, 30 min free, no card |
| Rosetta Stone | Immersion lessons and TruAccent scoring; free conversation isn't the core format | Self-guided software | around 25 | around $12–14/month, or a lifetime plan near $180 |
| Duolingo (app) | No: occasional speech-recognition prompts | 100% AI | 40+ | free, Super around $7–13/month |
| Duolingo Max | Yes: Video Call, free-form AI conversation | 100% AI | Video Call up to Japanese and Korean | around $30/month |
| Babbel | Limited: "Babbel Speak" AI feature in beta | Self-guided | 14, no Japanese or Korean | around $13/month, cheaper annually |
| Busuu | Yes, through guided scenarios | Self-guided, community corrections | 14, Japanese and Korean included | free, then around $9–14/month |
| Pimsleur | Yes, through scripted audio repetition | Self-guided audio method | 50+ | around $20/month |
| Preply / iTalki | Yes, with a human | Tutor marketplaces | 90+ / 150+ | $15–25/hour, or $4–20 a lesson |
Read this table in two passes. First the daily spoken conversation column: it separates the tools that make you produce speech from the ones that make you recognize it. Then the human-or-AI column: it drives the price, the availability, and whether you need to book ahead.
With Volpiko, you don't open an app: it calls you, at the time you picked.
The coach calls like an actual contact would, with their face on the screen and two buttons. You pick up, you talk for ten minutes, you hang up. They know your progress and the topics you've told them you care about. It's the same trick behind Rosetta Stone's daily streak, applied to speaking instead of matching: the tool comes to you instead of waiting for your motivation. The format is detailed on the calls page.
Ten minutes of conversation isn't nothing. That's dozens of sentences you produced yourself, in a language you didn't prepare in advance. Do it five days a week. Three months in, your relationship to speaking up has changed.
The rest of the product follows the same logic. Pronunciation gets scored sentence by sentence, not at the end of a unit. Short lessons run from pre-A1 to C2 in Spanish, with spaced repetition and a running count of words you've learned and words you've actually said out loud.
Sixty coaches are available, ten per language. Each one has a job, a city, an age, and interests, and they're reachable any time, by voice or by text. You can switch to someone else when a conversation goes stale.
One thing worth saying plainly: the coach is an AI. No human teacher steps in, not for individuals, not for companies. That's exactly what makes a daily call at 7:10 a.m. possible for €9.90 a month.
Volpiko has 6 languages open today; the bigger catalogs run into the dozens.
Eighteen languages are on the roadmap, six are open. If you're after Portuguese, Dutch, or Greek, look elsewhere without hesitation. Rosetta Stone covers around 25 languages, including Japanese and Korean. Preply lists 90+, iTalki 150+, with actual human tutors behind them. That's a different business, and it's a solid one.
Rosetta Stone itself keeps advantages Volpiko isn't trying to cover:
Duolingo Max also deserves an honest mention: its Video Call feature really does offer free-form AI conversation, Japanese and Korean included. Volpiko's difference is the format, not the existence of voice: an outbound call every day, a product built entirely around speaking, and a lower price.
And if you want a real human on the other end specifically for Spanish, Preply and iTalki do exactly that, often within a day of signing up. For a wider view of the field, see the best apps to speak Spanish in 2026.
If your goal is a strong foundation, a trustworthy accent, and a method you'll actually stick with, Rosetta Stone is a solid choice; if you want to talk, pick a tool where you talk.
Choose Rosetta Stone for a translation-free method, real pronunciation feedback, and the option to pay once instead of subscribing forever. Choose Preply or iTalki for a rare language, or for tailored human coaching. Choose Duolingo for the free tier, the gamification, and the widest catalog around.
Choose Volpiko if the block is in speaking and you want to talk from week one. A call a day, ten minutes, free conversation, pronunciation scored sentence by sentence. Thirty minutes of conversation are free at signup, no card required, and five minutes are open even without an account. After that, it's €9.90 a month for three hours of conversation, no commitment: see pricing.
Nothing forces you to pick just one, either. An immersion lesson in the morning and a call in the evening work well together. But if your budget only covers one subscription, go back to the question at the start. In six months, would you rather have matched Spanish, or spoken it. The answer names the tool.
Its immersion lessons train vocabulary, sentence patterns, and pronunciation through matching and repetition, and TruAccent scores you against native speakers in real time. Free, unscripted conversation isn't the core format. To produce spoken language, you need speaking time: someone asking you something you didn't prepare for and waiting on your answer. Check the publisher's official page, offers evolve.
The Common European Framework of Reference, published by the Council of Europe, separates reception, production, and interaction. A matching-and-repetition lesson mostly trains the first two: you recognize, you repeat, you get corrected. A conversation trains the third: you search for words live, correct yourself, answer without preparation. The two complement each other; neither replaces the other.
Ten to fifteen minutes a day beats one hour a week, because recalling a word in under a second is trained through closely spaced repetition, not through long, infrequent sessions. What matters is that the time is spent producing speech, not just listening to it. A short session, kept up five days out of seven, changes the reflex within a term.
It depends on the catalog. Busuu, Duolingo, Preply, iTalki, Pimsleur, and Rosetta Stone all teach Japanese and Korean; Babbel doesn't include either in its 14 languages. Volpiko has both open too, with pronunciation practice built in, since the point is speaking, not reading kana or hangeul. Always check whether your target language gets the same spoken-practice depth as the flagship one, it isn't automatic.
A human tutor on a marketplace runs around $15 to $25 an hour on Preply's listed rates, and dedicated one-on-one subscriptions elsewhere cost more. On the AI side, Duolingo Max runs around $30 a month and Rosetta Stone around $12 to $14 a month, or near $180 for a lifetime plan. Volpiko is €9.90 a month for three hours of conversation, no commitment. These prices move, verify them before you subscribe.
Volpiko
On réveille ton coach. Encore une seconde, et tu parles.