Youper AI Review 2026: The App Is Shutting Down. Here's What It Was and Where to Go Next.

Youper is closing. As of September 30, 2026, the app stops working, every account is deleted, and the AI mental-health chatbot built on cognitive behavioral therapy will no longer exist. Youper, Inc. confirmed the shutdown directly on its own site: subscriptions are already disabled, no further charges will occur, and anyone who wants their data has until that date to export it. If you found this page searching "is Youper worth it" or "is Youper shutting down," the second question now answers the first — this review covers what the app actually did, what the research behind it actually found, and what its users said was missing, so you know what you're replacing rather than just where to replace it.

What Youper Actually Was

Youper described itself as a "Wellbeing AI chatbot" — conversational software that tracked mood, walked users through cognitive behavioral therapy (CBT) exercises, and organized reflections into a running journal. It was co-founded in 2016 by psychiatrist Dr. Jose Hamilton Vargas, alongside Diego Dotta and Thiago Marafon, with the explicit aim of making evidence-based mental health tools cheaper and more available than a therapist's office allows.

The core loop was a daily "Quick Chat": a short, structured conversation in which the AI walked a user through a CBT thought-reframing exercise, logged the mood and topic to a Mental Health Record, and periodically surfaced an "Insight" — a pattern noticed across several days of entries. Under the hood, Youper drew on four separate therapeutic frameworks: CBT, Dialectical Behavior Therapy (DBT), Acceptance and Commitment Therapy (ACT), and Mindfulness-Based Cognitive Therapy, rotating between them depending on what the conversation surfaced. That is a genuinely broad clinical toolkit for a chat app to carry — most competitors pick one framework and stay there.

Explore: cognitive restructuring · dialectical behavior therapy

The Research Behind It — Read Honestly

Youper's most cited credential is a 2021 study in the Journal of Medical Internet Research, coauthored with two Stanford psychology researchers, examining 4,517 paying users who consented to have their data studied. The results were real: anxiety symptoms (measured by GAD-7) dropped with a Cohen's d of 0.57 over the first two weeks and 0.60 over four weeks; depression symptoms (PHQ) dropped with d of 0.46 and 0.42 over the same windows. Those are moderate effect sizes — not nothing.

Read the study's own limitations section and the picture gets more specific, which is what a study's limitations section is for. It was observational, with no control group, so there is no way to separate "the app helped" from "people who stick with a daily check-in for two weeks tend to feel somewhat better regardless of what the check-in is." The outcome measures were both self-report. Four of the six authors — Andrea Niles, Jose Hamilton Vargas, Thiago Marafon, and Diego Dotta Couto — are Youper employees and disclosed shareholders in the company; only the two Stanford-affiliated authors, Ashish Mehta and James Gross, declared no conflict. And the improvement curve flattens almost immediately: from day 14 to day 28, the anxiety effect size moved by 0.02 and the depression effect size actually ticked up slightly. Whatever the app was doing, most of it happened in the first two weeks and did not compound from there.

None of that makes the study worthless — a moderate, real effect in a large sample is genuine evidence, and industry-affiliated research is common across digital health. It does mean the honest label is "promising early-use signal from an industry-affiliated observational cohort," not "clinically proven," and a reader deciding whether any AI mental health tool is worth trusting should always check which of those two claims the marketing is actually making.

Explore: cognitive distortions

What It Cost

Youper sold one plan: $69.99 per year, with a 7-day free trial and no monthly option. There was no ongoing free tier — once the trial ended, the CBT chat and mood tracking required payment. As of the shutdown notice, that no longer matters in the way it once did: all auto-renewing subscriptions have been disabled, no new or renewing charges will occur, and anyone with prepaid time remaining is being sent refund instructions through their app store.

What Users Said Was Missing

The app's App Store rating held up — 4.8 stars across 15.3K ratings on iOS — while Google Play users were notably less satisfied, 3.9 stars across 49.4K ratings, a gap wide enough to suggest the experience varied meaningfully by platform. Long-time users specifically flagged a redesign that eliminated features they had relied on: meditation and journaling tools were cut, and multiple reviewers described the resulting app as "very limited" next to earlier versions. Others reported the app freezing during use. None of that is a knock on the underlying method — CBT, delivered by a chatbot or a therapist, is a well-studied approach — but it is a reminder that a subscription buys access to whatever the company decides to keep building, and that access can shrink or end entirely, which is exactly what has now happened.

Explore: journaling for clarity

Where a Continuous Coach Does Something a Daily Check-In Doesn't

Youper's Quick Chat was built to notice a pattern across entries — that is what the Mental Health Record and Insights features were for. But the unit of the product was the single daily check-in: a short CBT exercise, logged, then a fresh check-in tomorrow. It surfaced patterns; it did not carry a standing plan for you forward across weeks the way an ongoing coaching relationship does — checking whether last week's stated intention actually happened, not just whether this week's mood log looks different from last week's.

IX Coach is built around the opposite unit: not a check-in that resets, but a single AI coach that holds context across sessions and is explicitly designed to draw on twenty years of practice across psychological, coaching, and spiritual methods — not one framework rotated among several, applied by one continuous relationship rather than a fresh daily prompt. That is a real, checkable design difference, not a marketing claim: one product's core loop is a daily reset, the other's is designed around continuity. Whether that continuity matters more to you than Youper's narrower CBT-specific toolkit is a real tradeoff, not a foregone conclusion — someone who wants a tightly scoped, single-framework thought-reframing tool and nothing more may have been better served by Youper's approach while it existed.

Explore: values based action · self monitoring

Is an AI Mental Health App the Right Tool at All?

Youper, like most consumer mental health apps, was explicit that it was not therapy and not a substitute for one — the youper.ai site pointed users in crisis to 988 or the Crisis Text Line rather than attempting to handle acute risk itself. That is the right posture for this entire category of product, IX Coach included: none of these tools are a replacement for a licensed clinician when the need is clinical, and any app that implies otherwise should be treated skeptically.

Where an AI mental-health or coaching app fits is the space next to that: daily reflection, pattern-noticing, and structured practice of methods like cognitive restructuring or values-based action between the moments when you'd see a human professional, or in place of no support at all. Whether a chatbot or a continuous AI coach is the better format for that in-between space is a real, personal question — a well-designed daily check-in and a continuous coaching relationship both have a legitimate claim on it, and the choice mostly comes down to whether you want a tool you actively return to for a scoped exercise or one that carries the thread for you.

Explore: acceptance commitment therapy · mindfulness based cognitive therapy

Is Youper shutting down?

Yes. Youper, Inc. announced it is closing and winding down the business. The app and all accounts stop working on September 30, 2026, and every account, chat history, and piece of personal data is permanently deleted starting October 1, 2026.

Will I be charged if I keep my Youper subscription?

No. All auto-renewing subscriptions have been disabled — Youper states no future charges will occur. Anyone with prepaid time remaining is being sent refund instructions through their app store (Apple or Google Play).

How do I get my data out of Youper before it closes?

Through the app, go to Settings then Request My Data, or visit start.youper.ai/export-data if you no longer have app access. Data must be exported by September 30, 2026 — after that, Youper states it cannot be retrieved.

Was Youper clinically proven to work?

A 2021 study coauthored with two Stanford researchers found moderate reductions in anxiety and depression symptoms across 4,517 paying users over four weeks. It was observational with no control group, used self-report measures only, and four of its six authors were Youper employees and shareholders — real early-use evidence, not a randomized clinical trial.

What should I look for in a Youper alternative?

The same things worth checking on any mental health or coaching app before you trust it with something personal: what specific method it's built on, whether the evidence behind it is graded honestly (industry-funded and observational versus independent and controlled), what happens to your data if the company folds, and whether it's honest about not being a substitute for a licensed clinician in a crisis.

Frequently asked questions

Research

• Ashish Mehta, Andrea N. Niles, Jose Hamilton Vargas, Thiago Marafon, Diego Dotta Couto, James J. Gross, (2021), Acceptability and Effectiveness of Artificial Intelligence Therapy for Anxiety and Depression (Youper): Longitudinal Observational Study, Journal of Medical Internet Research, Vol 23, No 6 — Observational, no control group, self-report outcomes only, in 4,517 paying users; four of six authors are Youper employees and shareholders. Anxiety and depression symptoms dropped over the first two weeks (Cohen's d = 0.57 and 0.46) with almost no further change days 14 to 28 — real signal, industry-affiliated, and short-horizon, which is a different thing than a randomized trial of lasting benefit. verified 2026-08-27