Search for the best mental health app and you get listicles that rank a meditation timer, a teletherapy marketplace, and a mood diary against each other as though they were competing products. They are not. They do different jobs, and the reason so many downloads die inside a week is that people pick the wrong category rather than the wrong app.

The four categories, and what each is actually for

Almost everything marketed as a mental health app falls into one of four groups. Being clear about which one you need narrows a field of hundreds to a field of about five.

  • Therapy access — platforms that connect you to a licensed therapist by video or message. This is the only category that delivers clinical treatment, because the treatment is a human being.
  • Meditation and sleep — guided audio, breathing, sleep stories. Good for acute regulation: bringing your nervous system down in the moment.
  • Mood and symptom tracking — logging how you feel over time, usually as a chart. Useful as data, especially alongside therapy, but tracking is observation rather than intervention.
  • Daily reflection and structured self-help — guided prompts, CBT-style exercises, journaling with a shape to it. This is the category built for the long slow work of noticing and changing your own patterns.

Which one do you actually need?

A rough test: describe your problem in one sentence and notice its tense. If it is happening right now — I am spiralling, I cannot sleep tonight — you need the meditation and breathing category, because you need regulation in the next ten minutes. If it is chronic and clinical — I have not felt like myself for months, this is affecting my work and relationships — you need therapy access, and you need it before anything else on this list.

If it is recurring but not acute — the same argument keeps happening, I know intellectually this is not a catastrophe but I react as if it is — that is the reflection category. Recurring patterns need a repeated practice of noticing, not a one-off calming exercise. And if your problem is that you genuinely cannot tell what is going on, mood tracking first is reasonable: you cannot change a pattern you have not yet seen.

Why feature lists are a bad way to choose

Comparison articles rank by feature count, which is close to backwards. The variable that determines whether a mental health app helps you is whether you still open it in week six. Everything else is downstream of that. Retention is mostly a function of friction and honesty: how long the daily loop takes, whether it works on a bad day, and whether the app is straight with you about what it is.

  • How long is the daily commitment on your worst day, not your best? An app that needs twenty focused minutes is an app you will abandon during exactly the week you needed it.
  • Does it have a low-effort path? Good tools let you check in badly. Ones that demand a full entry or nothing train you to skip.
  • What happens when you break a streak? If the answer is guilt, the mechanic is working against the goal.
  • Is the free tier a real trial or a demo? You cannot evaluate a habit tool in five minutes; you need several real days.
  • Is it honest about not being therapy? Apps that blur this line are the ones to distrust generally.

The AI question

A growing number of apps now include an AI layer. The useful version reflects your own patterns back to you — noticing that your low-mood entries cluster on Sunday evenings, or that a worry you logged three weeks ago never actually materialised. That is a genuinely new capability, because no human, including you, reliably remembers across weeks.

The version to avoid is any AI that behaves as a therapist: interpreting, diagnosing, or advising on clinical matters. That is not a capability question, it is a safety one. An AI that will not tell you when something is beyond it is more dangerous than one with fewer features. Look for tools that route you to real help at the first sign of crisis rather than trying to handle it.

How to run a real trial

Pick one app in one category and use it for two weeks. Not three apps — one. Running several at once means you are managing tools instead of doing the work, and it makes it impossible to tell which one helped.

At the end of two weeks ask a single question: did opening it become easier or harder over time? Easier means the friction is right for you and it is worth continuing. Harder means the shape is wrong — usually too long, too demanding, or too performatively upbeat — and you should switch rather than conclude that self-help does not work for you. The category was probably right; the specific tool was not.

One last thing worth saying plainly: if you are in crisis, none of these are the answer. Contact a crisis line or emergency services. Every app in every category above is designed for the ordinary weeks, not the worst night.

This piece compares specific tools inside each category. If you haven't yet settled on which category you actually need, our companion guide on how to choose a mental health app walks through that first, faster decision. Mental health apps: how they work and how to choose one →