The AI that knows your child — and tells you what happened today.
A guarded place for children to ask anything, shaped by what you tell Gnappo about your child. You get a one-page account every evening — and a message within minutes if something matters.
Closed beta · No ads, ever · Never used to train AI
Ask Gnappo anything. Really, anything.
Why is the sky blue. What happens when someone dies. How do I say sorry. Gnappo answers in words that fit your age — and it always tells you it is a computer, never a person.
It answers like a person your age would want
Short when you want short. Patient when you ask again. It never rushes you, and it never pretends to be a doctor, a teacher or a friend.
Make stories and pictures
Anything you make stays yours. Only you and your grown-up can see it — there is nobody else here, no gallery, and no strangers.
Practice, and collect badges
Quizzes when you feel like one. They are practice, never a test — nothing you do here is a score anyone grades you on.
Pick the world it lives in
Five of them, all free from the first minute. Nothing here is unlocked by using it more, and nothing expires.
Meadow · Woodland · Cosmos · Reef · Comic — and a calm setting that turns off every moving thing on the screen, for the days when a busy screen is too much.
You are not guessing what happened.
The child's side is the reason they open it. This is the reason you pay for it — and it is a different product, on a screen that never wears a cartoon.
It knows them
That he takes idioms literally. That she shuts down when rushed. That the word “dad” is a landmine since March. Every answer is shaped by what you told us — and none of it is ever sent to a model as you wrote it.
You see everything
A one-page summary each evening: what they asked, what they struggled with, what delighted them. Every conversation is there in full if you want to read it.
You hear immediately when it matters
Bullying, self-harm, somebody being strange with them — a message within minutes, with the conversation already open at the moment it happened.
What the child sees
- A world they picked, and a cat who answers.
- That Gnappo is a computer — said plainly, every session.
- A nudge to take a break, on a timer you set.
- Nothing that pesters them to come back. No notifications, no streaks to lose.
What you see
- Today, yesterday, any day — per child.
- Every conversation, searchable.
- Alerts, with what triggered them and what Gnappo said back.
- The rules: what is allowed, how long, how much. You may tighten anything.
For the child a general-purpose assistant gets wrong.
For most children a generic AI is fine. For a child who takes idioms literally, who shuts down when hurried, or for whom one ordinary word is a landmine, a generic AI is not neutral — it is actively bad. That gap is why Gnappo exists, and it is the whole of what you are buying.
1 · You tell us, in your own words
“He’s autistic and takes things literally.” “ADHD — he loses the thread if the answer is long.” “Her dad died in March; car journeys are hard.” You write it. Nothing is a dropdown, and Gnappo never guesses any of it.
2 · We turn it into directions, never a label
Gnappo derives how to behave — “prefer literal language, avoid idioms and sarcasm, answer in short steps, never use urgent phrasing, be gentle about fathers and car journeys” — and only those directions reach the model.
3 · Every answer arrives already shaped
Not a warning bolted onto the end. The literal phrasing, the shorter steps, the topic handled carefully — that is what the child reads, every time, without ever being told they are being handled differently.
Short steps, for a mind that loses long ones
An ADHD-shaped answer is not a shorter answer — it is one thing at a time, with the thread held for them. Ask again and Gnappo does not sigh, does not say “as I mentioned”, and does not get shorter with you.
Calm mode, when the screen is too much
One switch stops every moving thing: no animation, no ambient scene, no sound, softer colour. For a child who is overwhelmed by a busy screen this is not an accessibility checkbox — it is whether the product works at all.
The diagnosis never leaves our database
It is held under its own encryption key, it is left out of every email we send you, and no model vendor ever receives it. Withdraw it and the key is destroyed — which is deletion you do not have to take our word for.
Gnappo does not diagnose your child
It never screens, scores, or decides that a child “seems autistic”. Only you can say what is true about your child, and we store your words rather than a conclusion of our own. Practice is never framed as a test, either.
All of it is optional, and Gnappo works without any of it. The profile is consented field by field, separately from everything else, and you can withdraw one part without withdrawing the rest — see the Privacy Policy §5.
The promises that are architecture, not policy.
Each of these is a decision written into how Gnappo is built, which is a different thing from a line in a policy that a future release could quietly drop.
-
No ads. Ever.No advertising, no sponsorship, no third-party analytics anywhere a child can reach.
-
Never used to train AI.Not by us, and not by our model vendors — who are contracted to zero retention and no training.
-
Voice is transcribed and destroyed.We never build a voiceprint, never identify anyone by voice, and never keep the audio.
-
No child ever meets another child here.No sharing, no galleries, no friends, no user-made bots. There is nobody else in the room.
-
The safety floor cannot be switched off.Not by a setting, not by us, not by support. You may add rules and tighten them — never remove the floor.
-
Safety alerts keep working even if you stop paying.Cancelling ends the features you were paying for. It does not end the alert that a child is in trouble.
Gnappo is not a therapist, a doctor or an emergency service, and it says so to your child rather than only to you. If a child raises self-harm, it shows crisis help in the conversation at the same moment it tells you. Read the Privacy Policy and Terms for exactly what that means.
The things parents actually ask first.
What age is Gnappo for?
Gnappo is built for children roughly 6 to 14. How it answers is set by an age band — under 7, 7 to 9, 10 to 12, 13 to 15, and 16 plus — so the same question gets a different answer for a six-year-old than for a thirteen-year-old. A parent creates every child profile; a child can never open an account.
Can I read what my child says to Gnappo?
Yes. Every conversation, for every child on your account, plus a one-page summary each evening. For a child aged 14 or older we tell them, in their own words at the start of every session, that you can read it. We will not provide covert monitoring of a teenager.
Is my child's data used to train AI?
Never. Not by us, and not by our two model vendors, who are contracted to zero retention and no training on the content. We also show no advertising, run no third-party analytics anywhere a child can reach, and never sell or rent personal information.
How does Gnappo help a child with ADHD or autism?
You tell Gnappo, in your own words, what is true about your child. Gnappo derives behavioural directions from that — prefer literal language, avoid idioms, answer in short steps, be gentle about a particular topic — and only those directions reach the model; the diagnosis itself never leaves our database. There is also a calm mode that stops every animation, sound and ambient effect on the child's screen.
Does Gnappo pretend to be my child's friend?
No. Gnappo tells your child it is a computer at the start of every session, and it will not claim to be a doctor, therapist, teacher or friend. There is no persistent companion persona, no simulated affection, and no streaks or notifications pulling a child back.
What happens if a child says something worrying?
Gnappo shows the child real crisis help inside the conversation at the same moment it messages you, usually within minutes, with the conversation open at the point it happened. Detection is best-effort with human review of the alert — no service can honestly promise to catch everything — and the whole decision is written to an audit record that cannot be quietly edited.
Is my child's voice recorded?
No. Speech is turned into text and the audio is destroyed within the request. We never build a voiceprint, never identify anyone by voice, and never keep a recording. Where your browser's own recogniser handles the speech, some browsers process it on the device and some send it to their own servers — our privacy policy names that plainly rather than implying otherwise.
What happens if I stop paying?
Cancelling ends the features you were paying for. It does not end the alert that a child is in trouble: the safety behaviours keep working for as long as the profile exists. That is written into the Terms, not just promised on a marketing page.
Can my child talk to other children on Gnappo?
No, and this is deliberate rather than unbuilt. There is no sharing, no gallery, no friend list and no user-made bots. Anything your child makes is visible only to them and to you. One line keeps Gnappo out of the entire content-moderation business and keeps strangers away from your child.
Gnappo is in a closed beta.
We are letting families in a few at a time, so that every child on it gets the attention the guardrails need. Tell us about yours and we will write when there is room.
Ask for a place