IX Coach

Mental Health Chatbot vs AI Coach: Two Products Wearing One Description

What is the difference between a mental health chatbot and an AI coach?

A mental health chatbot and an AI coach look identical in an app store. Both are conversations with software, both talk about stress and mood, both cost about the same. The line between them is not tone or quality — it is what each product claims it can do about a condition, which is the same line United States regulators use to separate a wellness product from a medical device. This page explains where that line sits, what each category is structurally built for, what happened when the best-known chatbot in the category retired its consumer app, and how to tell which one answers what you are actually dealing with.

A mental health chatbot and an AI coach are sold in the same aisle, described in almost the same words, and priced within a few dollars of each other. The difference that matters is not how warm either one sounds. It is what each product says it can do about a condition — and that single sentence, more than any feature either one ships, decides which rules apply to it, what evidence it has to produce before it can make its claim, and how stable it is likely to be a year from now.

Three products show the spread. Wysa is live in the App Store, where it describes itself as self-help techniques drawn from CBT and DBT for "everyday wellbeing challenges like low mood, stress, anxiety, sleep." Woebot, the most recognized name the category ever had, retired its consumer app on June 30, 2025; Woebot Health still trades, and now sells to providers and payers rather than to you. IX Coach sits on the far side of the line from both: it makes no claim about any condition at all, and the methods it coaches from are published as readable pages, so you can open something like approach versus avoidance goals and inspect the mechanism before you ever start a conversation. Same search result, three genuinely different objects.

Explore: approach vs avoidance goals

The line is what a product claims it can treat

United States regulators do not sort these products by how clinical they sound. They sort them by intended use. The FDA's general wellness guidance, issued January 2026, puts it in one sentence: software intended "for maintaining or encouraging a healthy lifestyle and is unrelated to the diagnosis, cure, mitigation, prevention, or treatment of a disease or condition" is not a device.

Read that twice, because it inverts the assumption most people bring to this comparison. Clinical vocabulary is not what makes something regulated — a claim is. An app can say the letters CBT, teach a recognized technique, and remain a general wellness product, so long as it does not claim to treat the condition that technique was developed for. The moment it does claim that, it becomes a medical device and has to earn an authorization before it can say so.

Which means the honest version of this comparison is uncomfortable: a great many products marketed as mental health chatbots and a great many marketed as AI coaches are, in regulatory terms, the same kind of object. The category names are marketing. The claim is the thing with teeth.

What a mental health chatbot is structurally good at

A chatbot in this category is built around a library. It holds a set of techniques with decades of research behind them, and it delivers one to you on demand, at two in the morning, without an appointment and without your name attached. That is a real thing to be good at, and nothing on the coaching side of the line does it better.

Wysa is the clearest live example. Its App Store listing carries a 4.8 rating across 24K ratings, is published by Touchkin eServices Private Limited, and prices its tiers from $9.99 to $144.99 depending on what you buy. Its own description leads with anonymity — "Your identity will remain anonymous and your conversations are privacy protected" — which is a design choice, not a slogan, and it is exactly why some people will talk to a chatbot about something they have never said to a person.

One detail from that same listing is worth more than any comparison table: among Wysa's in-app purchases is a $29.99 line for "1:1 Life Coaching Sessions." A product filed under mental health chatbot sells coaching inside itself. If you were hoping the two categories divide cleanly, they do not — not even within a single app.

Explore: acceptance commitment therapy

What an AI coach is structurally good at

A coaching product is built around a direction rather than a library. Its whole shape assumes you have named something you want to move toward and that the work is closing the distance. This is not a softer version of the same job — psychologists have a name for the distinction, approach versus avoidance goals, and the two produce measurably different behavior. A tool built to reduce an unpleasant state and a tool built to advance toward a chosen one are optimizing different things, and you can feel which one you are talking to within a few exchanges.

That is also where the structural weakness sits. A coaching product assumes you can name the thing. On a week when you cannot, a technique library asks less of you and gives you somewhere to start, and it is the better tool that week.

Where IX Coach differs from most of the coaching side is checkable rather than promotional: it does not route you through a fixed track keyed to a mood score. It works from what you actually said, and every method it draws on has a page you can read — the GROW model, committed action — which means you can audit the coaching before you take any of it. Whether that produces better conversations for you is a question only your own use answers.

Explore: approach vs avoidance goals · the grow model · committed action · the coaching habit

Woebot retired its consumer app, and the category did not stay still

Woebot was the reference implementation of a mental health chatbot for most of a decade. Its consumer app was retired on June 30, 2025. The company is still operating: its site today describes the work as building "the future of chat-based AI wellness tools," and markets to providers and payers rather than to individual users.

Several outlets have attributed the retirement to the cost of pursuing FDA marketing authorization. That explanation is plausible and widely repeated, but the primary and freely readable sources reachable for this page state the date and not the reason — so this page does not assert one. The sequence is the checkable part, and the sequence is the part that matters to a reader: the most established product in this category stopped selling to consumers, and the people using it had to choose again.

That is the risk profile of the category, stated plainly. A product resting on a claim about a condition is exposed to whatever a regulator decides about that claim. A product that never made the claim is not.

Explore: choice overload

The regulator has not finished deciding

On November 6, 2025, the FDA's Digital Health Advisory Committee met on "generative artificial intelligence-enabled digital mental health medical devices." The committee considered, in the agency's own wording, "the benefits, risks to health, and risk mitigations that might be considered for these new digital mental health devices, including premarket evidence and postmarket monitoring considerations."

A committee convened to discuss what the rules should be is a committee telling you the rules are not settled. For anyone choosing a product this year, that is practical information rather than trivia: the part of this market that makes clinical claims is being actively reshaped, and the part that does not is not waiting on the outcome.

If you need help right now

This is a reference page about software. It is not a crisis service, and neither is any chatbot, coach, or app named on it — including IX Coach. If you are in distress, or you are frightened for someone else, reach a person now. These lines are free, they are staffed by people, they are available in the United States, and they do not require you to be using anything described here.

Call or text 988 — the 988 Suicide & Crisis Lifeline, free and confidential, 24 hours a day.

Text HOME to 741741 — the Crisis Text Line, if talking out loud is not possible right now.

Call 911 — if someone is in immediate physical danger.

Where a person beats both, and it is not close

If what you are carrying needs assessment, a diagnosis, or a treatment plan, that is a licensed clinician's work and no product on either side of this line does it — not the chatbot using clinical vocabulary, not the coach, and not IX Coach. A software product that is honest about its category will tell you this, and one that blurs it is telling you something about itself.

A human coach beats the coaching apps on a narrower but real axis too: a person remembers you across months without being reminded, notices the thing you did not say, and can be held accountable by a professional body. What software offers against that is availability and price, which for a lot of people is the difference between some support and none. That is a genuine trade, not a marketing one, and it should be made with open eyes.

Choosing between them without over-researching it

The research on choosing suggests that past a certain point, more comparison makes decisions worse rather than better — the pattern behind satisficing versus maximizing, and behind choice overload. This is a low-stakes, reversible, roughly $10-to-$30 decision that people routinely treat as a high-stakes one.

Choose a mental health chatbot if you want a technique on demand and anonymity matters more to you than continuity — if the useful question is "give me something to do with this feeling right now."

Choose an AI coach, IX Coach included, if you can name where you are trying to get to and the useful question is "what is actually in the way" — and if being able to read the method behind the conversation matters to you.

Choose a person if the honest answer is that you need to be assessed rather than supported. Both categories of software are cheap enough to test in a week, and a week of your own experience will tell you more than another evening of comparison pages, this one included.

Explore: satisficing vs maximizing · choice overload · choice architecture

Is an AI coach just a mental health chatbot with different marketing?

Sometimes, and the only reliable way to tell is to read what the product claims. Under the FDA's general wellness guidance, a product that does not claim to diagnose, treat, or mitigate a condition is not a medical device regardless of what it calls itself. So two apps in different app-store categories can be the same regulatory object. What genuinely differs is design intent: a chatbot organizes around delivering a technique, a coach organizes around moving toward a goal you named.

Are mental health chatbots regulated by the FDA?

It depends entirely on what each one claims. A product that stays within general wellness — supporting a healthy lifestyle, unrelated to treating a condition — is not regulated as a device. A product that claims to treat or mitigate a psychiatric condition is a medical device and needs authorization. The FDA's Digital Health Advisory Committee met on November 6, 2025 specifically to consider how generative-AI products in this space should be evaluated, which indicates the framework is still being worked out.

Why did Woebot's app disappear?

Woebot Health retired its consumer app on June 30, 2025. The company continues to operate and now sells to providers and payers, describing its work as chat-based AI wellness tools. The reason for the retirement has been widely attributed to the cost of pursuing FDA authorization, but the sources readable for this page state the date without stating a cause, so no cause is claimed here.

Frequently asked questions

What is the difference between a mental health chatbot and an AI coach?

A mental health chatbot and an AI coach look identical in an app store. Both are conversations with software, both talk about stress and mood, both cost about the same. The line between them is not tone or quality — it is what each product claims it can do about a condition, which is the same line United States regulators use to separate a wellness product from a medical device. This page explains where that line sits, what each category is structurally built for, what happened when the best-known chatbot in the category retired its consumer app, and how to tell which one answers what you are actually dealing with.

Is an AI coach just a mental health chatbot with different marketing?

Sometimes, and the only reliable way to tell is to read what the product claims. Under the FDA's general wellness guidance, a product that does not claim to diagnose, treat, or mitigate a condition is not a medical device regardless of what it calls itself. So two apps in different app-store categories can be the same regulatory object. What genuinely differs is design intent: a chatbot organizes around delivering a technique, a coach organizes around moving toward a goal you named.

Are mental health chatbots regulated by the FDA?

It depends entirely on what each one claims. A product that stays within general wellness — supporting a healthy lifestyle, unrelated to treating a condition — is not regulated as a device. A product that claims to treat or mitigate a psychiatric condition is a medical device and needs authorization. The FDA's Digital Health Advisory Committee met on November 6, 2025 specifically to consider how generative-AI products in this space should be evaluated, which indicates the framework is still being worked out.

Why did Woebot's app disappear?

Woebot Health retired its consumer app on June 30, 2025. The company continues to operate and now sells to providers and payers, describing its work as chat-based AI wellness tools. The reason for the retirement has been widely attributed to the cost of pursuing FDA authorization, but the sources readable for this page state the date without stating a cause, so no cause is claimed here.

Research

Practice this with IX Coach

Try this practice

Keep reading