Personalized iPhone keyboard · Beta waitlist open

It knows what
you meant.

MEYNT is an iPhone keyboard that studies how you type: where your thumbs actually land, the names and slang you use every day, and the corrections you keep undoing. Every message teaches it a little more, so autocorrect stops guessing and starts getting it right.

  • No setup or training required
  • Learning stays on your iPhone by default
  • We never sell typing data
The problem

Autocorrect shouldn't fight you.

Every keyboard ships with one model for millions of people. But nobody types the same way. You hold your phone differently, hit keys slightly off-center, use your own slang and your friends' names, and you end up correcting the keyboard instead of the other way around.

Your words get "fixed"

Names, nicknames, inside jokes and slang are treated like typos, again and again.

see you at Starbuck → you meant Starbucks

Your thumbs land off-target

If you always hit the right edge of a key, a one-size model reads that as the neighboring letter.

I'll try → you meant tell her later

Undoing never teaches it

You reject the same correction for the tenth time and the keyboard makes it an eleventh.

lol LOL → undone 10× this week
The shift

MEYNT learns you.

Instead of one generic model, MEYNT builds a personal typing profile from four signals you already produce every time you type. Nothing extra to do: just keep texting.

01

Learns your taps

MEYNT maps where your fingers naturally land on each key, so a tap that drifts toward the edge still becomes the letter you were aiming for.

Signal: tap position
02

Learns your words

Names, slang, abbreviations, and the way you spell things become part of your personal vocabulary instead of being flagged as mistakes.

Signal: personal vocabulary
03

Learns from corrections

When you keep rejecting the same autocorrect, MEYNT treats that as feedback and stops making the change.

Signal: accept / reject history
04

Understands context

Grammar, the words around your cursor, and common phrases help MEYNT pick the suggestion that actually fits the sentence.

Signal: language context
The idea in 10 seconds

Watch MEYNT learn you.

One message, typed three times. The first time, a generic correction gets it wrong. Then MEYNT picks up the signals. The next time, the same typing produces the right result.

Before MEYNT learns
running late, where are we meeting?
Start typing…
Wrong correction
Traditional autocorrect

One model for everyone

  • Same dictionary and key targets for every user
  • Treats unfamiliar names and slang as errors
  • Repeats corrections you've already undone
  • Learning happens out of sight, so you can't see or shape it
MEYNT

A personal typing model

  • Key targets reshaped around where your thumbs land
  • Your vocabulary is learned, not corrected
  • Rejected corrections become a signal to stop
  • A visible profile showing what it has learned (coming soon)
Privacy, built in

A keyboard sees everything you type. We take that seriously.

Personalization only works if you trust it. MEYNT is designed to do its learning on your iPhone, collect only what the feature needs, and give you the controls to see, reset, or delete it.

On-device firstYour typing profile is built and stored on your phone by default.
No message archiveMEYNT learns patterns, not transcripts of your conversations.
Never soldWe don't sell or rent typing data, and we don't use it for ads.
You're in controlReset learning, remove words, or delete everything at any time.
Roadmap

Where MEYNT is headed.

We're building in the open. Here's what's working today, what testers will see next, and what's planned after launch. Timelines can change as we learn from beta testers.

Built

Working prototype

  • Custom iOS keyboard
  • Personalized autocorrect engine
  • Tap-position learning
Next

Private beta

  • Waitlist testers via TestFlight
  • Correction memory
  • Personal vocabulary tuning
Planned

My MEYNT

  • See what the keyboard has learned
  • Typing profile and key heatmaps
  • Adaptive typing test
Planned

Beyond launch

  • App Store release
  • Additional languages
  • Optional encrypted backup

Ready for a keyboard that adapts to you?

Join the waitlist to get early access, beta-testing invites, and the occasional behind-the-scenes update from the team. No spam. Unsubscribe anytime.

Home / How It Works
How it works

Your keyboard gets to know how you type.

MEYNT personalizes from everyday typing. In the background, it builds a model of how your thumbs move, what you write, and which corrections you refuse. Then it uses that model to rank every suggestion it makes. Here's what happens, step by step.

01

You type normally

Install the keyboard, switch to it, and keep texting the way you always do. There's no calibration screen and no required training.

Day one feels like a familiar, capable keyboard.
02

MEYNT notices patterns

As you type, it records lightweight signals: where each tap lands on a key, words you use often, corrections you accept or undo, and the context around them.

Signals are processed on your iPhone.
03

Your profile takes shape

Those signals combine into a personal typing profile: adjusted key targets, a personal dictionary, and a memory of what you don't want changed.

The profile is yours. Reset it anytime.
04

Suggestions improve

Autocorrect and predictions start reflecting you. Fewer wrong fixes, fewer retypes, and suggestions that sound like the way you write.

The more you type, the sharper it gets.
The four signals

Four signals. One personal model.

Each signal answers a different question about what you meant. MEYNT weighs them together before deciding whether to correct, suggest, or leave your word alone.

Tap learning

If you consistently hit the right side of T, MEYNT learns that a tap landing between T and Y is still more likely to mean T for you. Over time, each key gets an invisible target shaped around your thumb, not an average person's.

What's kept: per-key tap offsets (x/y positions relative to the key), not the text you typed.

Personal vocabulary

Friends' names, your school or team, slang, abbreviations, and uncommon spellings gradually join your personal dictionary. Once a word is yours, MEYNT suggests it instead of "fixing" it.

What's kept: individual words you use repeatedly, with how often you use them.

Correction memory

If MEYNT changes something and you repeatedly change it back, that's evidence. After enough rejections, it stops making that correction. Corrections you accept are reinforced.

What's kept: pairs like "lol → LOL: rejected 6 times", not the message they appeared in.

Language intelligence

MEYNT also considers spelling rules, grammar, word frequency, the words just before your cursor, and common phrases. That's how it can tell "their", "there" and "they're" apart in a sentence.

What's kept: nothing extra. Context is read in the moment to rank suggestions.

Try tap learning

Tap the T key, the way you naturally would.

Each dot is one tap. The lilac ring is where MEYNT thinks the center of your T is. Tap a few times near one edge and watch the target move toward your habit. This is a simplified version of what happens on each key of the real keyboard.

0taps on T
0 pxlearned shift
0%confidence

Tap T about 10 times to build a confident profile.

What to expect

How personalization builds over time.

There's no finish line. MEYNT keeps adjusting as your habits change. This is a typical progression; how fast it goes depends on how much you type.

First hour

A solid baseline

Standard autocorrect and predictions. MEYNT starts collecting tap positions in the background.

First few days

Your words stick

Names and slang you use repeatedly stop being corrected. Rejected corrections start to fade out.

First couple of weeks

Keys fit your thumbs

With enough taps per key, targets shift toward where you actually hit. Near-misses resolve correctly.

Ongoing

It keeps up with you

New phone, new case, new slang? The profile adapts gradually so it stays accurate.

Optional

Want it to learn faster? Take the typing test.

A short typing exercise, about two minutes, gives MEYNT a concentrated set of taps across every key. It's a head start, never a requirement. Everyday typing is enough to start learning. In a future update, the test will adapt to focus on the letters where your profile has the least data.

Home / Features
Features

Built around how you actually type.

Everything in MEYNT serves one goal: fewer wrong corrections and suggestions that sound like you. Features marked "Coming soon" are planned and in design, not yet in the beta. We label them clearly so you always know what you're getting.

Beta

Personalized autocorrect

Corrections are ranked using your own typing profile instead of a model that treats every user the same. The result is fewer "fixes" you have to undo.

Why it matters: autocorrect should save you time, not create a second job.

Beta

Tap position learning

Builds a per-key model from normal typing, so taps that consistently drift left, right, up or down still register as the letter you meant.

Why it matters: big thumbs, small screens, one-handed typing and phone cases all change where you hit.

Beta

Context-aware suggestions

Uses grammar, the previous words in your sentence, and common phrases to avoid random or awkward suggestions in the prediction bar.

Why it matters: "see you their" should never be the top suggestion.

Beta

Personal words

Learns names, slang, abbreviations and repeated vocabulary automatically. Once a word is part of your dictionary, MEYNT suggests and completes it.

Why it matters: your best friend's name isn't a typo.

Beta

Correction memory

Remembers corrections you repeatedly accept or reject. Undo the same change enough times and MEYNT stops making it.

Why it matters: you should only have to say no a few times.

Beta

Privacy-focused learning

Personalization is designed to happen and stay on your device. No ads, no selling data, and clear in-app explanations before any permission is requested.

Why it matters: a keyboard has to earn trust before it can learn anything.

Coming soon

My MEYNT

A dashboard in the MEYNT app showing what the keyboard has learned about you: your personal words, corrections it now skips, and how your profile is improving.

Why it matters: learning you can see is learning you can trust and fix.

Coming soon

Typing profile

A visual heatmap of your learned key patterns, showing where you tend to tap on each key and which keys still need more data.

Why it matters: it turns an invisible model into something you understand.

Coming soon

Adaptive typing test

An optional test that targets the letters and combinations where the model has the least confidence, so a couple of minutes fills the biggest gaps.

Why it matters: faster personalization without typing random sentences.

Coming soon

Word controls

Add, edit or remove personal words yourself, and tell MEYNT to never correct a specific word, on top of the automatic learning.

Why it matters: sometimes you want to teach it directly.

Planned

More languages

Support for typing in additional languages, including switching between languages in the same conversation.

Why it matters: lots of people don't text in just one language.

Planned

Optional encrypted backup

An opt-in way to keep your typing profile when you get a new iPhone. Off by default, and described fully before you turn it on.

Why it matters: you shouldn't have to start from zero on a new phone.

Side by side

What changes when your keyboard learns you.

SituationTypical keyboardMEYNT
You type a friend's unusual nameOften flagged or replaced until learnedAdded to your personal words after repeated use
You always tap the right edge of a keyKey targets are the same for everyoneKey target shifts toward your habit
You undo the same correction againMay keep making itRejections are remembered; the correction stops
You want to know what it learnedLearning is mostly hiddenMy MEYNT shows your profile (coming soon)
You want a fresh startReset buried in system settingsOne-tap reset inside the MEYNT app
Home / Privacy Policy
Legal

Privacy Policy

A keyboard can see almost everything you type, so we've written this policy to be specific. It explains what MEYNT collects, why, where it's stored, who can access it, how long we keep it, and the choices you have.

This policy describes how MEYNT is designed to work during development and beta. Before public launch, it will be reviewed by a parent, guardian or qualified legal professional and updated to match the final product. If anything in the app ever works differently from what's written here, this policy will be updated before that change ships.

Privacy at a glance

Learns on your deviceYour typing profile is built and stored on your iPhone by default.
Collects only what's neededTap positions, repeated words and correction choices, used only to personalize.
You control itTurn learning off, reset your profile, or request deletion anytime.
No message archiveWe don't store the full text of your conversations.
No selling or adsWe don't sell, rent or share typing data for advertising.
No passwordsiOS blocks third-party keyboards from password and other secure fields.
Home / Terms of Use
Legal

Terms of Use

These terms explain the rules for using the MEYNT website, waitlist, beta and keyboard. They're written to be readable. Please take a few minutes with them.

These terms will be reviewed by a parent, guardian or qualified legal professional before MEYNT's public launch, and may be updated as a result.

Home / About
About

Why we built MEYNT.

MEYNT began with a simple frustration every phone user knows: fighting your own keyboard. Here's how that turned into a product and a team.

The problem

We were correcting the keyboard.

Autocorrect is supposed to make typing easier. But for many of us it felt like the opposite: retyping names, undoing the same "fix" over and over, and sending messages that said something we never meant. The keyboard was confident, and often wrong.

The realization

Nobody types the same way.

We noticed that each of us made different "mistakes". One person always clipped the edge of the space bar. Another hit Y when they meant T. Everyone had their own words. If people type differently, their keyboards shouldn't all behave exactly the same.

The vision

A keyboard that understands the person behind the typing.

MEYNT's goal is a keyboard that learns your taps, your vocabulary, your corrections and your context, and shows you what it has learned. The name says it: it should know what you meant.

Today

From prototype to team.

MEYNT started as a student project and grew into a team covering product, design, outreach, privacy and operations, developed through programs like LaunchX. We have a working iOS keyboard prototype and are preparing for a private beta with people from our waitlist.

What we believe

Three principles behind every decision.

Personal beats generic

The best suggestion is the one that fits the person typing, not the average user. We measure success by how rarely you have to correct MEYNT.

Trust comes first

A keyboard sees everything. That's why we keep learning on-device, collect the minimum, explain permissions before asking, and never sell data.

Learning should be visible

You should be able to see what your keyboard learned, fix it, and reset it. Hidden models are hard to trust; visible ones are yours.

Home / Team
Team

A student team with different specialties.

Built for LaunchX and beyond. Every role on MEYNT exists because of something that person actually ships, from the keyboard engine to the privacy policy you can read on this site.

M

Muhsin

Product & Engineering

Builds and leads development of the MEYNT keyboard and iPhone app.

  • Autocorrect and personalization engine
  • Typing-data learning models
  • Product testing and iteration

Achievements

  • Built MEYNT’s working native iOS keyboard and personalization system
  • One of the youngest students to complete the Code Ninjas Black Belt project
  • Competitive U18 pickleball player, ranked 35th worldwide
  • Honors Algebra 2 student as a 9th grader
  • Robotics team member (CAD)
  • JV tennis player as a freshman
  • Competitive soccer player for 5 years

Shipped: the first working iOS keyboard prototype.

Ai

Aiden

Web Platform, Privacy & Legal Lead

Owns MEYNT's public presence and the trust framework behind it, making sure every promise MEYNT makes about your data is accurate, clear, and matches how the keyboard actually works.

  • Leads the MEYNT website and waitlist experience
  • Leads the Privacy Policy and Terms of Use
  • Turns technical data flows into plain-language disclosures
  • Structures creator and partnership agreements

Achievements

  • Built and launched MEYNT’s website
  • MEYNT’s Lead Editor & Operations Manager: oversaw team logistics, organized project information, and led final quality verification
  • 4.0 GPA while taking 3 AP classes as a 9th grader
  • AP Precalculus and AP Chemistry student as a 9th grader
  • Junior Thespians National Qualifier
  • Seasoned debater
  • UIL Science team competitor
  • Science Fair competitor
  • Competitive JV soccer player

Shipped: a 17-section privacy policy and full terms, aligned line by line with product behavior.

D

Dhruv

Brand & Design

Leads MEYNT's visual identity across every surface people see.

  • Logo and app icon
  • Typography and website style
  • Social media branding

Achievements

  • Created MEYNT’s app logo, color scheme, and design
  • Placed top 20 out of 78 contestants in a hackathon
  • Honors Algebra 2 student as a freshman
  • Made the freshman basketball team at competitive Coppell High School
  • Black belt in Kukkiwon Taekwondo
  • First place in multiple YMCA swim meet events

Shipped: the MEYNT visual system, end to end.

A

Aarav

Partnerships & Outreach

Handles how MEYNT talks to the world and the businesses around it.

  • Company communication
  • Sponsorship and brand partnerships
  • Waitlist strategy and local businesses

Shipped: outreach to 25 potential sponsors.

T

Tyler

Hardware & Operations

Researches what it would take to bring MEYNT beyond software.

  • Hardware concepts and manufacturing costs
  • Pricing and minimum order quantities
  • Outreach tracking and operations

Achievements

Shipped: a map of manufacturing costs and MOQ paths.

+

Open role

Growth & Social Media

We're looking for someone to grow MEYNT's audience.

  • Content distribution and social analytics
  • Creator outreach
  • Turning views into testers and waitlist signups

Interested? Email hello@meynt.app with "Growth" in the subject.

Home / FAQ
FAQ

Questions people actually ask.

Short answers about what MEYNT is, how it learns, what happens to your data, and when you can try it. Can't find yours? Email hello@meynt.app.

No questions match that search. Email hello@meynt.app and we'll answer it.

Still curious?

The fastest way to see MEYNT is to try it. Join the waitlist and tell us you'd like to beta test.

Home / Join the Waitlist
Join the waitlist

Be one of the first to try MEYNT.

Sign up for launch news and a chance to join the private beta. It takes about 30 seconds, and we'll only email you about MEYNT.

Please enter your first name.
We'll send launch news and beta invites here. Please enter a valid email address, like you@email.com.
Helps us test on the phones our users actually have. MEYNT is iPhone-only for now.
Interested in beta testing? Optional
Please confirm the age and privacy checkbox to continue.

We'll only use your email for MEYNT updates and beta opportunities. No spam, never sold, unsubscribe anytime.

What happens next

  1. You're on the listYou'll get a short confirmation email from hello@meynt.app.
  2. Beta invites roll outWe invite testers in small groups through Apple's TestFlight, prioritizing people who said yes to beta testing.
  3. You help shape MEYNTTesters get a quick feedback channel and early looks at new features like My MEYNT.
  4. Launch dayEveryone on the list hears first when MEYNT arrives on the App Store.

Follow our progress

Behind-the-scenes builds, design drops, and demos of MEYNT learning in real time.

Home / Contact
Contact

Say hello.

Pick the inbox that fits your question so it reaches the right person. If you're not sure, hello@meynt.app works for everything and we'll route it from there.

Generalhello@meynt.app

Questions about MEYNT, the waitlist, press and media requests, or joining the team.

Replies in 2–3 business days
Supportsupport@meynt.app

Help with installing the keyboard, Full Access, the beta, or anything that isn't working. Opens with the beta.

Include your iPhone model and iOS version
Partnershipspartnerships@meynt.app

Brands, sponsors, creators and local businesses interested in working with MEYNT.

Replies in about a week
Privacyprivacy@meynt.app

Data access or deletion requests, and questions about our Privacy Policy. Please email from the address you signed up with.

Requests handled within 30 days