The GLP Fish guide

A course of treatment, kept properly.

How to set up a course, log doses and costs, weigh in at any hour and still compare fairly, read a forecast you can check the sources of, and pause without losing anything.

Written for the app as it is, with what shipped in 1.0 and what is in development labelled as such.

What GLP Fish is

GLP Fish is an iPhone app for one job: keeping a readable record of a GLP-1 course of treatment. It covers semaglutide and tirzepatide in their approved forms — the weekly injection and the daily tablet — from the weeks you are still deciding, through every dose and weigh-in, into a pause or a maintenance stage, and back to the same record if you resume.

It has three tabs. Records holds the whole timeline — weight, doses, costs, meals and body measurements, in the order they happened, across every stage. Today is where you log things quickly. Peers is an optional community feed. The app is fully usable without an account; signing in syncs your record across your devices.

Every number the app shows about expected weight loss comes from a published clinical study or published real-world data, and the source is named next to the number. What the app never does is give medical advice.

Read this first

GLP Fish is not a medical device. It does not provide medical advice, diagnose conditions, treat side effects, or recommend dose changes. Always follow your doctor's instructions on taking your medication, adjusting your dose, or stopping treatment. Research scenarios show study-group averages and do not represent your individual results.

Setting up a course

First launch takes a few minutes and asks only for what the app needs.

The basics

GLP Fish asks for your birth year, height, current weight and target weight, whether you have diabetes, and where you are with medication: still considering, already using, or paused. Nothing here leaves your iPhone unless you create an account.

If you are already on treatment

Give the date your current course really started — and the pause date, if you are paused. Earlier weights can be back-filled later on the Records tab, so starting the app late does not mean an empty record.

Choosing your medication

Three medications are covered: semaglutide weekly injection (Wegovy), semaglutide daily tablet, and tirzepatide weekly injection (Zepbound, sold as Mounjaro in some markets). Official medication information is matched to your market — FDA, PMDA, EMA, MFDS, TFDA — and opens inside the app. Dose lists are the delivered dose in milligrams, for example 0.25 to 2.4 mg for the semaglutide injection, never the volume inside a multi-dose pen.

Dose schedule and reminders

Set your current dose and a fixed day and time — weekly for injections, daily for the tablet. Reminders are local notifications on your iPhone, created only for the times you confirm, and the old ones are cancelled whenever your plan changes. The app keeps two things apart: the plan you and your prescriber actually agreed on, and any comparison you save while still deciding.

Costs

Enter what you actually pay — per dose or per month — in your own currency: US dollar, Japanese yen, euro, Korean won, Chinese yuan, New Taiwan dollar, or Hong Kong dollar. Every logged dose can carry its own cost, so a price change is recorded when it happens, not averaged away.

Weigh-ins and the correction

You can weigh in at any hour. The app corrects for when and how you weighed, so readings compare fairly.

Each entry stores three things: the number on the scale, the time, and the context you pick — morning before food, after a meal, before dinner, bedtime, and so on. The raw number you entered is never rewritten; the correction is a derived value shown beside it.

How the correction works

All comparisons are made against the morning, fasted reading — the steadiest point of the day. When you log a reading in another context, the app subtracts what that context typically adds: an after-dinner reading tends to sit higher than the same body would fasted the next morning, so it gets a larger correction than a mid-afternoon one. Every corrected reading carries a confidence range that says how much the correction is worth — narrow in the morning, wide late in the day — and the range is shown with the number, not hidden.

These context averages describe everybody at once, and they are deliberately rough starting points rather than a validated formula. The app says so by showing its uncertainty instead of one tidy number.

Learning your own rhythm (1.1, in development)

The update now in development goes further. Once you log a morning reading and a later one on the same day a few times, the app starts estimating your own daily swing and blends it into the correction, with the confidence range narrowing as it learns. It narrows to a floor, never to a single point — a body keeps some noise, and the chart should not pretend otherwise.

What a reading is compared against

A new reading is compared with a similar one — same context, similar time of day — and the result says both what you entered and what it was compared against. A weight that held steady is recorded as a result in its own right, because holding steady through a pause is exactly what you wanted to know.

Three habits worth keeping

Log the context honestly. The correction can only be as good as what you tell it.

Keep weighing in at odd hours. A morning and an evening reading on the same day is exactly the pair that teaches the app your rhythm.

Let one reading be one reading. The trend is the record; a single point, high or low, is noise the app is built to absorb.

The forecast and its sources

GLP Fish never invents a weight curve. Every number about expected loss comes from a published study, named on screen next to it.

Before you start: planning scenarios

Pick the medications you are considering and a horizon — 12, 24, 52 or 72 weeks — and the app draws each study curve from your starting weight. The anchors are the trials the app cites: STEP 1, in which the semaglutide 2.4 mg weekly injection group averaged −14.9% body weight over 68 weeks (the placebo group averaged −2.4%); SURMOUNT-1, in which the tirzepatide 10 mg weekly injection group averaged −19.5% over 72 weeks (placebo −3.1%); and OASIS 4, in which the semaglutide 25 mg daily tablet group averaged −13.6% over 64 weeks. Each curve names its study, and the study population is described beside it.

The placebo arm is kept visible on purpose. Everyone in those trials also had diet and activity counselling, and reading the drug arm's number as the drug's doing alone would credit the drug with work the participants did.

These curves appear when the study evidence plausibly covers you. If your recorded profile falls outside the trial population, the published endpoints stay readable as text, and the app does not draw a personalised-looking curve it cannot stand behind.

Once you are on treatment

The app does not restart a study curve from your latest weight. It shows your recorded trend beside the published endpoints, so you can see where you stand against the study group without the app pretending the trial was run on you.

The prediction card (1.1, in development)

The update now in development replaces several separate exhibits with one prediction: an expected weight change, a range around it, and the share of people who reached at least a given level. Its anchor is real-world rather than trial data — a published study by Gasoyan and colleagues (Obesity, 2025) of 7,881 adults who started injectable semaglutide or tirzepatide in 2021–2023, in which the one-year average was −7.7% on semaglutide and −12.4% on tirzepatide, and −11.9% for the people still on treatment at a year.

Why real-world numbers and not the trials? Trials have dedicated staff, fixed dose increases and high adherence; in the real world about half the people had stopped within a year. Real life is where the app's readers live. The prediction also adjusts for your starting weight — the same drug takes a smaller percentage off a heavier body, a finding the regulators' own reviews spell out — and your own records can move it, slowly and within limits, because a few good weeks are not a new biology.

What the forecast is not

Group averages, twice over: the trial curves are what a study group averaged, and the real-world prediction is what a real population averaged. Neither is a promise about you, and neither recommends a dose change. Deciding what to do with what you see belongs to you and your prescriber; the app's job is that you walk into that conversation with a record.

Logging doses and costs

From the Today tab, a dose takes seconds to log — and the record stays honest.

An injection entry records the drug, the dose in milligrams, the time you took it, an optional injection site, and what it cost you. A tablet entry records the dose and time the same way. The two stay separate — shots and pills follow different rules and keep different records.

A dose is logged as scheduled or as taken late. The original schedule and the actual time are both kept, so an honest record never looks like perfect adherence — and your timeline shows what really happened.

The missed-dose window

The app shows the official rule for your chosen medication, straight from the label. For the semaglutide weekly injection: if more than 48 hours remain before your next scheduled day, take it when you notice, then return to your fixed day. For the tirzepatide weekly injection: within 96 hours after the missed day, keeping at least 72 hours between any two doses. For the daily tablet: skip the missed dose and take the next one the next day — never a double dose. Outside the window, or after several missed doses in a row, the app does not invent a new schedule; it tells you to contact your prescriber.

Costs by stage

Every dose can carry a cost in your currency, and each treatment episode keeps its own total, so a pause and a resume read as two honest chapters instead of one blurred line. In the planning comparison you can set assumptions per dose or per month and see totals over the horizon you choose.

Pauses, resumes and maintenance

Pausing is a structured step in GLP Fish, and nothing is erased.

A pause records the date and a reason you pick: goal reached, side effects, cost, availability, your clinician's advice, a planned break, or something else. The reason goes into the record because a year from now, "why did I stop" deserves an answer made of facts.

The finished episode becomes read-only, but it stays on the timeline, and the chart marks each stage as its own chapter with its boundaries. You keep weighing in and logging meals through the pause — whether the weight holds without the medication is exactly the question the pause exists to answer.

Resuming is a dedicated step too: you re-select the medication and dose, confirm the current prescription, and a new episode starts on top of the old one. Maintenance is the same record viewed from the other side — a stage of its own on the same timeline, tracked the same way as everything else.

Meals, measurements and Peers

Weight is the headline; two more kinds of record sit beside it.

Meals are logged next to your weight, with a protein target shown as a range of roughly 1.0 to 1.2 grams per kilogram of body weight. A day you did not log is treated as unlogged — the app does not assume you ate nothing.

Body measurements — chest, waist, hips and body fat — are each recorded with their own date, and a figure only combines them when they were taken in one sitting. The app does not splice a Tuesday waist onto a Friday weight.

The Peers tab shows public posts from other readers, tagged with language and region. Posting, liking and commenting needs an account. Every post can be reported to moderation, and the update now in development extends reporting to comments and adds author blocking, with the blocked list managed from your profile.

Your data and privacy

A health record is sensitive, and GLP Fish treats it that way.

Without an account, everything you record stays on your iPhone. The app is fully usable that way, including the forecast and every stage of the treatment record.

An account is a username and a password — nothing else is required, and an email address is optional, used only for recovery if you choose to add one. Sign in and your record syncs across your devices. Delete the account and the cloud copy goes with it.

The first time you sign in on a device that already has local records, you choose: import them into the account, or keep them separate. An account that already has data never has a shared device's guest records mixed into it.

If the cloud copy cannot be read safely, the app refuses to load it rather than overwrite what is on the device. There are no advertising SDKs in the app, and health records are not sold or shared with third parties.

The 1.1 update, now in development, adds a read-only Apple Health import for weight readings, with your explicit consent. The app never writes to Apple Health.

Read the GLP Fish privacy policy

Tips and troubleshooting

The questions we expect most, answered the way the app actually behaves.

Why does the same body read differently at different hours?

Food, water, salt and clothing all move the number, and the morning fasted reading is the steadiest one. GLP Fish corrects each reading by its context and shows the confidence range of that correction, so an evening reading can be compared with a morning one without pretending they were taken the same way.

Weeks of records, and the forecast has barely moved. Why?

The planning curves never move — they are the studies' numbers, not yours. Your own records can move the prediction being introduced in the 1.1 update, but slowly and within limits: the app weights a few weeks lightly against a year of published data, so a fast start or a slow month does not rewrite your outlook.

I missed my injection. What does the app do?

It shows the official window from your medication's label — 48 hours for the semaglutide weekly injection, 96 hours with at least 72 hours between doses for tirzepatide. Inside the window it tells you to take the dose and return to your fixed day, and you log the dose as taken late. Outside the window it tells you to contact your prescriber, and it does not reschedule on its own.

My reminders did not arrive.

Reminders are local notifications on your iPhone, created only for the fixed times you confirmed in your plan. Check that iOS allows notifications for GLP Fish, and check that the day and time in your plan are the ones you wanted. Changing the plan cancels and recreates the reminders.

I typed the wrong weight, or logged a dose I did not take.

Records are editable on the Records tab: a weight can be corrected with its context and time, and an entry that should not exist can be deleted. One boundary is deliberate — an injection date cannot be moved before the start of its treatment episode.

Do I need an account?

No. Guest records live on the device and every feature works on them. An account adds sync across devices, the Peers feed, and deletion of the cloud copy when you delete the account. When you first sign in, you decide whether local records are imported or kept apart.

My medication is not in the list.

GLP Fish covers the approved forms of semaglutide and tirzepatide: the weekly injections and the daily tablet, with official label information for your market. Compounded and other unapproved products are deliberately left out — regulators have warned about their quality and dosing, and a record built on them would be a record of the wrong thing.

Will the app tell me to change my dose?

No. Dose lists, missed-dose windows and study curves are educational record-keeping. Whether a higher dose is tolerable or wise is a decision for your prescriber with your full history — which is exactly the conversation your record is for.

GLP Fish

On the App Store

Start the record on day one

GLP Fish is coming to the App Store for iPhone. Follow the link to see the listing.

View on the App Store

iPhone · English, 简体中文, 繁體中文, 日本語, Français, Español, 한국어

GLP Fish is not a medical device. It does not provide medical advice, diagnose conditions, treat side effects, or recommend dose changes. Always follow your doctor's instructions on taking your medication, adjusting your dose, or stopping treatment. Research scenarios show study-group averages and do not represent your individual results.