Italian flashcards, and the arithmetic nobody prints

Here's the honest version of what you're buying, before anything else on this page.

Flashcards build recognition quickly and production slowly. Run a deck for two months and you'll understand Italian you couldn't understand before. You will still stall mid-sentence reaching for a word you "know," because recognizing magari on a card and producing it at speed in a conversation are different skills, and cards train the first one hard and the second one barely.

So they're a component of learning Italian rather than a method. Worth doing, worth 15 minutes a day, not worth expecting a conversation from.

I'll say that out loud because I build a flashcard app and it costs me something to say. The three pages ranking above this one don't say it, and the reason is structural: all three compete on how much material they have.

Flashcardo publishes the 1,000 most common Italian words in 20 blocks of 50, 18 topic decks, 20 YouTube video decks and 8 printable PDFs. StudyStack lists 100 Italian sets on its first page alone, across roughly 11 pages, from a 7-card set to one with 1,581. Digital Dialects has 16 sets of 11 to 17 words each, with audio on every card. All three read on 21 September 2026.

Not one of them tells you how long any of it takes.

What flashcards do to your Italian, and what they don't

The format works, and the evidence is better than most product pages bother to cite.

Retrieval beats restudying, by a lot. Karpicke and Roediger taught college students 40 Swahili-English word pairs and tested them a week later. Where pairs kept being retrieved after the first correct answer, students recalled about 80% of the list. Where pairs were dropped from testing and restudied instead, 36% and 33%.

Spacing beats cramming. Cepeda and colleagues reviewed 317 experiments on distributed practice. Same total minutes, spread across days, buys more memory than the same minutes in one sitting.

Karpicke & Roediger 2008 · one week later

0

% recalled when pairs kept being tested after the first correct answer

0

% recalled when pairs were restudied instead and dropped from testing

That last detail is the one worth carrying around. The learners could not tell. Which is a reason to distrust your own sense of whether a deck is doing anything, in either direction.

Now the limit, stated as precisely as the research allows.

Kang, Gollan and Pashler had learners see pictures of objects and hear their Hebrew names. One group repeated each name aloud; the other tried to produce it from memory first, then heard it. Two days later the retrieval group was better at producing the words out loud, with no loss of pronunciation quality. So retrieval practice does transfer to speech.

But that test was naming single pictures in isolation. It is not evidence that you'll produce the word inside a live sentence at conversational speed, and I've never found a study that shows a deck alone gets you there.

Webb's work on repetition gives the other half of the honest picture: even at 10 encounters with a word in context, "to develop full knowledge of a word more than ten repetitions may be needed." A card is one kind of encounter. You need the others.

1,000 words, in days

This is the number all three pages leave blank, and it's the one that decides whether you finish.

Your daily review load comes from your new-card rate, never from the size of the catalog. Anki's own manual states the ratio plainly: "If you are consistently learning 20 new cards a day, you can expect your daily reviews to be roughly about 200 cards/day." The same manual warns that more than one user "excitedly studied hundreds of new cards over their first few days of using the program, and then has become overwhelmed by the reviews required."

Roughly 10 reviews a day for every new card you add. Hold that ratio against Flashcardo's 1,000 Italian words.

The middle row is the one I'd pick. 10 a day finishes Flashcardo's whole frequency set in about 3 months on roughly 10 minutes a day, and 10 minutes survives a bad week in a way 20 does not.

The arithmetic in the last column is mine and it's rough: 6 seconds a card is a brisk pace, and your real number is worth measuring rather than assuming. I've written up how to pick your own daily number from your actual seconds-per-review, including why lapses push the true cost per card closer to 13 reviews than 10.

Set that against what 1,000 words buys you, because the catalogs imply it's a finish line.

Nation (2006) found that unassisted comprehension of spoken text at 98% coverage takes 6,000 to 7,000 word families, and written text 8,000 to 9,000. His description of what the weaker 95% coverage feels like is the line I think about most: roughly 7 unknown words in every minute of speech at a normal 150 words a minute.

So 1,000 words is a real, useful floor, and Italian's top of the curve genuinely carries ordinary conversation further than its length suggests, which is how frequency distributions work. It is not a destination, and a page that hands you 1,000 cards without saying so has set you up to feel like you failed.

One piece of good news specific to Italian. The Foreign Service Institute puts Italian in Category I at 24 weeks of class, the bottom of the range, where Spanish and French both sit at 30. That's the State Department's estimate for diplomats hitting professional working proficiency on 23 class hours a week, so it's not your timeline. But Italian is at the cheap end of the cheap category, and the six-week gap has real causes.

Frequency is the right order until it isn't

Flashcardo's 1,000 words come from Pinhok Language's Italian Frequency Dictionary, and the page credits the source, which I respect. Ordering by corpus frequency is the best sequencing logic on this entire search.

It beats the alternative, which is grouping by topic, and which nearly everything else here does. StudyStack's most-starred sets include Italian Colors at 12 cards and Parts of the Body at 42. Digital Dialects runs 11 colours in one set and 17 birds in another.

Learn 11 colours in one sitting and you don't get 11 words. You get one blurry colour-shaped slot with 11 candidates in it, and for weeks you'll pull azzurro when you wanted verde. Tinkham measured the penalty in 1993 and again in 1997, Waring replicated it in 1997, and Erten and Tekin found the same thing in young learners in 2008. Nation's 2000 guideline is the practical takeaway: teach one, let it settle, then the next.

The effect isn't unanimous. Higa (1963) found no penalty for antonym pairs, and Nakata and Suzuki (2019) found spacing itself shrinks the interference, so a decent scheduler buys back some of what a bad order costs.

Frequency order dodges most of that by accident, since related words rarely sit adjacent in a frequency list. It has its own blind spot, and Italian hands you a clean example: conoscere and sapere both mean "to know" and both live high in the frequency band, so a raw list will introduce them within days of each other. Hold them apart deliberately. Same for essere and stare.

The deeper limit arrives later. Past the first couple of thousand words, frequency stops predicting what you need, because what you need stops being average. The words I lose in Italian conversation are the ones attached to what I talk about, and no corpus knows that. Every deck on this search is somebody else's list: Pinhok's dictionary, a teacher's upload of Ciao! Sixth Edition Chapter 1, a set of 17 birds.

What goes on one Italian card

All three pages describe their cards. Here's one instead, so you can argue with the real thing.

Three choices in there have sources behind them, and one has a caveat the marketing pages skip.

The blank. Producing a word beats spotting it in a lineup, which is the retrieval finding above. It's why the front is a gap rather than a flip. Flashcardo's and StudyStack's cards are bare English-Italian pairs, so the strongest version of the format isn't available on either.

The audio. A word you can read and cannot say is half a word, and Bassetti and Atkinson (2015) found the written form still bends pronunciation in learners with years of instruction. Digital Dialects is the only one of the three with audio at all, and it says so in its first sentence, which is to its credit. Ours is synthesised speech generated per card with a native-accent instruction. Babbel records real people across its content, and on that specific axis Babbel beats both of us.

The image, with a caveat. The picture-on-a-card claim gets overstated by everyone selling cards, including people in my position. Carpenter and Olson (2012) tested exactly this on foreign vocabulary and found pictures did not initially beat plain translations, and made learners more overconfident about the words they'd seen with a picture. The benefit appeared only once that overconfidence was corrected by retrieval practice or an explicit warning.

One Italian-specific thing worth adding, which none of the three does. Put the article on the card with the noun. Il problema is masculine despite the ending, la mano is feminine despite hers, and if you learn the bare noun you'll be guessing gender out loud for a year. The article is part of the word you'll actually say.

Phrases deserve card status too. Ci vuole, mi raccomando, fa niente: these do more conversational work than most single words, and a word list structurally can't contain them.

What all of this costs

This is where the search fails buyers hardest, so I'll print everything I could confirm on a first-party page on 21 September 2026.

Flashcardo is free for the web decks, the 1,000-word frequency set, the 8 printable PDFs and the YouTube decks. Its Kindle ebook has a Buy Now button and no price anywhere on the page. Its spaced-repetition tool sits behind Register, with no statement of what the algorithm does or what it costs once you're in.

StudyStack is free to use. "Upgrade to remove ads" appears three times above the fold with no figure attached, and the page never says what else the upgrade changes.

Digital Dialects is free, with no commerce on the page at all.

Anki is free on Windows, macOS, Linux and Android, syncing through AnkiWeb is free, and iOS costs US$24.99 once rather than as a subscription.

Verbamor, which I build, is $19.99 a month, or $129.99 a year on the annual plan, which works out at about $10.83 a month. The 7-day free trial is on the annual plan only. 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. There's no free tier, and I'm not going to pretend there's a way to try it without a card.

Three of those are free and one of them is $130 a year, so let me address that rather than hope you don't do the subtraction.

The money doesn't buy a bigger catalog. Flashcardo's frequency set is good and free and I'd tell you to start there. What it covers is the labour of making cards, which is where every flashcard habit actually dies. Gabriel Wyner, who has built more language cards by hand than almost anyone, puts card creation at 2 to 3 hours a week to keep a 25-a-day habit running. That's the line item.

The mechanism is that you record a lesson or a conversation, or paste a transcript, and the app pulls out the words and phrases you actually reached for and fumbled, then builds them into cloze cards with audio and an image, scheduled by FSRS. Grammar moments come from your own corrected sentences the same way. A tutor helps and isn't required: the Translate tab saves single words and whole phrases through the same pipeline, so you can build cards with no lesson at all.

The limits, which matter more than the features. iOS only, so no Android and no browser. It holds what your lesson or your reading already gave you and teaches you nothing new, which means a true beginner should build their first thousand Italian words somewhere else first. And it covers 5 languages rather than working as a general-purpose flashcard tool.

Go use one of the free ones if this is you

Three groups should close this page, and I mean it literally.

Starting Italian from zero? Use Flashcardo's 1,000 most common words, free, and come back in three months. Frequency order is the correct answer for your first block of vocabulary, and we don't teach you words you've never met. Our whole design assumes you already met the word somewhere.

Enjoy making cards? Use Anki. The engine underneath it is better than what most paid products wrap around theirs, and it costs nothing outside iOS. Switch FSRS on yourself at the bottom of the deck options page: Anki still ships with SM-2 from 1987 as its default, whatever the blog posts claim. It comes with no Italian in it at all, which is the trade. You supply every word.

Studying for a test on a fixed list? Use StudyStack or a printable PDF. If your Ciao! Chapter 3 vocabulary quiz is on Thursday, somebody has already uploaded that exact set, and spaced repetition aims at remembering in 6 months, which is the wrong objective for a quiz with a date on it.

The case for paying starts where you're having real conversations in Italian, losing specific words out of them, and finding that no catalog contains the ones you keep losing.

One more thing about the free decks, which I'd want to know as a buyer. StudyStack's sets are user-submitted and unvetted, so quality swings hard: three different sets on page one are titled "214 cards," dated 2008, 2022 and 2024, and two near-identical "Classroom Greetings & Expressions" sets sit at 35 and 34 cards a year apart. And Digital Dialects' two Tell the Time sets still promise to teach you the time "in Indonesian," left over from another language, on a page whose footer reads 2023. Free is a fair price for that. Just know what you're getting.

Questions people actually ask

---

Sources

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