Anki Pro vs Anki

INTRODUCTION

Search "Anki" on the iOS App Store and at least three apps show up. One has an orange star icon. Another has a blue logo. A third might call itself something slightly different but still use the word "Anki" in its name. For a first-year medical student trying to download the tool half their class already uses, this is a genuine problem. Anki Pro vs Anki is not a choice between a free version and a premium upgrade. These are two completely separate products built by different people, running different code, using different algorithms, with different business models. The original Anki has existed since 2006. Anki Pro appeared fifteen years later, built by a different company, after its creator was explicitly asked not to. One is free and open-source. The other charges a subscription. One runs the most advanced scheduling algorithm in the field. The other still uses code from 1987. This is the full story of both, why the confusion exists, and what it actually means for anyone trying to learn with flashcards.

The Original: Anki Since 2006

Anki was created by Australian developer Damien Elmes. The first version appeared on 5 October 2006. Elmes built it to help himself learn Japanese and released it as free, open-source software under the AGPLv3 license. That means anyone can read the source code, modify it, or build on top of it.

Twenty years later, the ecosystem looks like this. Anki Desktop runs for free on Windows, macOS, and Linux. AnkiDroid is free on Android with over ten million Google Play downloads. AnkiWeb provides free cloud sync across all devices. The only paid component in the entire ecosystem is AnkiMobile for iOS, priced at $24.99 as a one-time purchase. That single revenue source funds everything else: the desktop app, the sync servers, the add-on infrastructure, and Elmes's full-time work maintaining it all.

The community built around Anki is enormous. Over 1,600 add-ons extend the software in ways Elmes never anticipated. The AnKing Step Deck alone has been downloaded more than 300,000 times. A 2024 survey found that 86.2 percent of US medical students use Anki, with 66.5 percent reviewing cards daily. No other flashcard tool comes close to that adoption rate in medical education.

In February 2026, Elmes announced he is gradually transitioning Anki's business operations and open-source stewardship to AnkiHub, founded by Nick Flint (known online as The AnKing) and Andrew Sanchez. Elmes cited burnout from years of solo maintenance. The public commitments: Anki stays open-source, no venture capital, no price increases, and AnkiDroid remains independently governed.

The Copy: How Anki Pro Was Built

Anki Pro appeared on the iOS App Store around late June 2021. It was developed by Vedas Apps Ltd, a London-based company. The Android package name is com.vedasapps.flashcards. The person behind it, according to Anki's creator, is Maksim Abramchuck.

Here is where the story gets unusual. Abramchuck actually emailed Elmes before launching to ask permission. Elmes responded directly. He asked Abramchuck not to build a competing app using the Anki name. He explained that AnkiMobile sales are the sole funding source for Anki's development and servers, and that a competing app trading on the Anki name would put the future of the entire free ecosystem at risk. Abramchuck replied with reassurances that he did not want to harm the Anki ecosystem.

Then he built and released the app anyway.

Elmes posted the exchange publicly on the Anki Forums on July 31, 2021. He noted that the new app copied answer button designs, used screenshots from the Image Occlusion add-on and AnkiMobile marketing materials, and appeared to redistribute shared decks from AnkiWeb against their license terms. Anki's official FAQ page now lists Anki Pro alongside AnkiApp as apps that "were developed by separate groups of people" and "are not related to the rest of the Anki ecosystem." The page states directly: we suspect the names were deliberately chosen to take advantage of the brand recognition we have built up.[link]

Under trademark pressure (Anki is a registered trademark of Ankitects Pty Ltd), Anki Pro rebranded to Noji on 2 June 2025. The announcement framed it as a "refresh." The timing matched legal enforcement. The URL slug noji.io/ankipro-vs-anki still targets the old brand name for search traffic.

The Algorithm Gap: FSRS vs SM-2

The technical difference between these two apps runs deeper than the interface. Understanding it requires knowing what happened inside Anki in October 2023.

For seventeen years, Anki used SM-2 as its default scheduling algorithm. SM-2 was designed by Piotr Wozniak in 1987 for SuperMemo. It assigns each card an "ease factor" starting at 2.5. When a student answers correctly, the interval before the next review is multiplied by this factor. When they struggle, the factor drops. The problem: the factor drops fast and recovers slowly. After several failures, cards get stuck in what the Anki community calls "ease hell," where intervals barely grow regardless of how many times the student answers correctly.

In October 2023, Anki version 23.10 introduced FSRS (Free Spaced Repetition Scheduler) as a native alternative. FSRS was developed by Jarrett Ye and the Open Spaced Repetition community. Instead of one ease factor, FSRS models memory through three variables: stability (how strong the memory trace is), difficulty (how hard the card is for this specific learner), and retrievability (the probability of successful recall at any given moment). The model was trained on roughly 727 million reviews collected from about 10,000 Anki users, the largest spaced repetition dataset ever assembled.

Benchmarks on this dataset show FSRS needs approximately twenty to thirty percent fewer reviews than SM-2 to achieve the same retention level. That is not a marketing claim. It is a measured result from a published KDD 2022 paper and subsequent open-source benchmarking.

Noji still runs on SM-2. It offers preset configurations labeled "Study languages" and "Study medical material," but the underlying engine has not adopted FSRS or any published machine-learning scheduler. That means the app charges a subscription for access to the older, less efficient algorithm. Meanwhile, Anki offers FSRS for free.

Why the Confusion Works So Well

The confusion between Anki and Anki Pro is not accidental. It follows a pattern that cognitive science research on brand recognition would predict: when two products share a name, consumers default to the one that appears more polished or more familiar in their app store search results.

Anki Pro marketed itself with a modern interface, bright design, and the word "Pro" in its name, which implies an upgraded version. For a student who has heard classmates mention "Anki" but has never installed it, the blue-logoed "Anki Pro" with a clean App Store listing looks more legitimate than the older, orange-starred app with a $24.99 price tag. The assumption is natural: the free one must be the basic version, and the "Pro" must be the real deal.

The scale of confusion is significant. Anki Pro / Noji has accumulated roughly 1.8 million Android downloads alone. App Store reviews are full of students who realized too late that they downloaded the wrong app. One reviewer wrote that they bought a subscription thinking it was the official Anki, only to discover their existing decks from AnkiWeb could not sync. Another described paying for a year before learning that "Anki Pro" was made by a completely different company.

Forum threads on the Anki community site show the same pattern repeatedly. Users post asking which app is the "correct" one. Others describe paying for a subscription and being unable to use decks they already built. Moderators respond with the same guidance: the real Anki apps are listed at apps.ankiweb.net and nowhere else. Anki Pro is not the only app that has traded on the name. A separate product called AnkiApp (now AlgoApp) created identical confusion years earlier using the same strategy.

The Science That Separates Them

The practical difference between these two apps is not just about branding or pricing. It connects to a body of research on memory that stretches back more than a century.

Hermann Ebbinghaus first documented the forgetting curve in 1885, showing that people lose roughly half of newly learned material within a day without review. A 2015 replication by Murre and Dros confirmed the core finding: memory decay follows a predictable mathematical pattern, and properly timed review interrupts that decay. The entire value of any spaced repetition algorithm is deciding when to show each card to prevent forgetting while minimizing unnecessary reviews.

SM-2 does this with fixed rules. FSRS does it with personalized predictions. Both work. But the efficiency gap is measurable, and it compounds over months. For a medical student reviewing 300 cards per day, twenty percent fewer reviews means sixty cards less per session at the same retention rate. Over a semester, that is hundreds of hours. The research on retrieval practice by Karpicke and Roediger (2008) confirmed that actively retrieving information from memory strengthens long-term retention more effectively than passive review. An efficient scheduling algorithm maximizes retrieval events while minimizing wasted repetitions.

Noji gives students SM-2 behind a subscription paywall. Anki gives students FSRS for free. Both claim to use "spaced repetition." Only one has the published research to back up the specific implementation.

How to Switch: The Migration Path

For anyone who downloaded Anki Pro by mistake and built card decks inside it, there is a way out. The community-built Copycat Importer add-on (code 2072125761, developed by Abdo) transfers card content and media files from Noji exports into real Anki. The process runs through Anki Desktop: open Tools, select Add-ons, click Get Add-ons, enter the code, restart, then import.

One important limitation: scheduling history does not transfer. Review intervals and ease factors built up inside Noji stay behind. Cards arrive in Anki as if they are new. For students deep into a study cycle, this means a temporary spike in review load as FSRS recalibrates each card's memory state from scratch. Starting fresh with a better algorithm is still worth it over the medium term.

CONCLUSION

Anki Pro and Anki are not versions of the same product. Anki is a twenty-year-old open-source project backed by a global community, over 1,600 add-ons, and a machine-learning scheduling algorithm trained on hundreds of millions of reviews. Anki Pro, now called Noji, is a four-year-old commercial app that was built after its creator was asked not to use the Anki name. It runs SM-2 behind a subscription, does not sync with AnkiWeb, and has no add-on ecosystem. The confusion is not a minor branding issue. Thousands of students have paid for subscriptions, built decks, and studied for months inside an app that does not connect to the ecosystem their classmates use. Knowing the difference before downloading matters. For those interested in how FSRS and SM-2 actually compare at the algorithm level, the technical details make the gap even clearer. The broader lesson is simpler: when two apps share a name, check who built them before trusting either with months of study data.