Three language apps still have notification permissions turned on across my phone, and only one of them gets opened without a push reminder doing the nagging for it.
Quick disclosure before the audit starts: some of the links below are affiliate links, and I earn a commission if you sign up through them, at no extra cost to you. Every app named here has lived on my phone or desktop for months at a stretch, not days, and I've cancelled more of them than I currently pay for.
Spaced repetition is the quiet mechanism behind most of this audit, whether an app admits it or not — Spaced Repetition Systems decide when a word resurfaces right before you'd otherwise forget it, rather than how cute the mascot is. My grandmother was Italian and never quite got around to teaching me the language, which is the plain, unglamorous reason Italian still gets more of my attention than anything else on this list. Spanish only rode along for a trip, and Japanese is the language I keep circling in my head without ever actually opening an app for it, which is its own kind of subscription math.
The Habit-Loop Apps: Mondly and the Streak Trap
Mondly by Pearson is the app built around habit-stacking, and a catalog of 41 languages means the Japanese I keep daydreaming about is sitting right there whenever I actually commit to it. The Daily Lesson ties to a streak counter that gets me through a drill most mornings before I open my invoicing software, though a streak and an actual habit are not quite the same thing — a distinction I've picked apart in more detail elsewhere on this site. The AR mode is a strange, genuinely useful party trick: a virtual instructor standing in my living room, which did more for spatial memory with new vocabulary than I expected going in.
Yara Kimani, a collaborator I've worked with on client design projects for years, still has an unbroken Mondly streak and mentions it with the quiet smugness of someone who has simply never missed a day. Translation-heavy drills are where the app starts showing its ceiling: recognizing a memorized phrase and actually generating new sentences from grammar rules are two different skills, and Mondly leans hard on the first one. For the longer version of how that stacks up against the other app most people default to, I wrote up Mondly vs Duolingo for long-term retention separately.
One-Time Audio Courses vs. Subscription Models
Switching my drills to Rocket Languages during walks around Olbrich Botanical Gardens turned out to be the bigger shift than any single app update this year. The interface looks like it hasn't been touched since 2018, and it skips levels, badges, and leveling-up sounds entirely, which turns out to be the point rather than a flaw: with no dopamine hit to lean on, the audio itself has to carry the entire lesson. The Pimsleur narrator has always felt like a slightly judgmental houseguest in my head, but the Rocket instructors feel more like someone actually invested in whether I survive an airport. Recording my own line and playing it back against a native speaker's version is the first tool that's actually shown me where my pronunciation drifts, rather than just marking an answer right or wrong, and that same playback loop is the closest thing to real speaking practice I've found without booking a tutor or a conversation partner.
It's a one-time purchase rather than a subscription, and that changes the math on whether it's worth it: no auto-renew to forget about, no monthly fee quietly running while I'm buried in client work, just a single payment and a 60-day money-back guarantee that's unusually generous for this category. The specials board at the Italian deli near my apartment is handwritten in cursive, and reading the whole thing out loud without reaching for my phone once is a small, concrete marker of what audio-first instruction is actually good for — training recall under real time pressure, not matching a translated phrase on a screen. For the longer version of that shift, I wrote about swapping streak-chasing for a real course library separately.
English Fluency Needs a Different Format Than a Vocabulary App
EF English Live sits in a different category entirely, built around live group classes with certified teachers rather than solo drills, plus private one-on-one lessons booked on demand when a specific meeting or presentation is bearing down on you. Group classes force actual back-and-forth timing and real-time response, which is a different kind of readiness than recognizing a phrase on a flashcard. That same real-time pressure is where most adult beginners report the sharpest anxiety, well before vocabulary gaps ever become the actual bottleneck. Certificates tied to CEFR levels matter if a course needs to go on a resume or satisfy an employer, which is a different goal than ordering dinner without panicking. None of this helps if the language in question isn't English — it's a tool built for one specific job, not a general-purpose app, and treating it like one is the fastest way to end up disappointed by it.
Matching the App to the Actual Problem, Not the Other Way Around
Spanish only entered the rotation because of a trip, and a deadline changes which app makes sense entirely — cramming survival phrases for a specific date rewards an app built for fast transactional phrases, not one built for long-term grammar architecture. Freelance work means my actual available minutes are scattered in short gaps between client calls rather than one quiet focused block, so anything that demands sustained attention gets abandoned no matter how good it is. Japanese stays a daydream rather than an active subscription because kanji is its own separate skill from spoken vocabulary, closer to learning a visual system than memorizing new words, and I haven't been honest with myself about whether I actually want to take that on. Italian is the one language where I've plateaued in an interesting way: comfortable enough to order food and read a menu, stuck enough that a real conversation still outruns me, which is a different problem than beginner apps are built to solve.
Auditing a subscription stack once a year, instead of letting it run quietly in the background, is the only part of this whole process I'd actually call a system. A reader named Fiona Ramirez, who has a habit of resurfacing in my inbox every time a post like this one goes up, emailed after an earlier Mondly review to say she'd cancelled hers that same week for the same translation-heavy reason, and we ended up comparing notes on Babbel, which neither of us had opened in months despite still being charged for it. Babbel is still technically active on my own card for a similar reason — cancelling felt like admitting defeat, and I stopped opening it after the third week of a subscription that then ran for half a year unused, which is a more honest explanation for most dead subscriptions than any actual flaw in the app itself.
Here's What Actually Stayed on My Card
Mondly stays for Japanese, specifically because a catalog of 41 languages means it's already paid for the moment I'm ready to start, and daily habit-stacking is genuinely the right mechanism for a brand new language with zero existing vocabulary. Rocket Languages stays for Italian, because grammar architecture and audio recall are what actually move an intermediate plateau, and a one-time cost with a real money-back window is a lower-risk bet than another monthly fee I'll forget to cancel. EF English Live only earns a place on this list for someone whose actual goal is English specifically, for work or a certificate, not for anyone casually picking up a third or fourth language — matching the tool to the specific goal matters more than any single app's feature list.
If you're staring at your own subscription graveyard, start with whichever one of these three actually matches the problem you have right now rather than the one with the flashiest onboarding screen: the audio-first course above is where I'd point someone stuck on a plateau, and the daily habit app above is where I'd point someone starting completely from zero. If there's one rule from a year of doing this: decide what a subscription needs to prove within the first month, write that down somewhere other than your own memory, and cancel anything that hasn't proven it by the date you picked — before the renewal makes that decision for you.