Capture
Say it however it comes out. “two eggs and toast” is enough. Or point the camera, scan a barcode, paste a link.
Launching soon on iPhone
Type “2 eggs, toast, and black coffee.” TypeCal turns it into a clean nutrition log, ready for your final review.Type “2 rotis, dal, and chai.” TypeCal turns it into a clean nutrition log, ready for your final review.
Join the waitlist for launch updates. Unsubscribe anytime.
Built for the days most trackers lose. Shared plates, late dinners, leftovers, quick bites — log the meal in plain language and keep the habit intact. Accuracy benchmark →
What changes
Less guessing. Fewer missed days. A calmer way to stay in range.
See the benchmarkShared plates, late dinners, leftovers — still logged.
“I log dinner while it is still on the table.”
How it works
Logging dinner should never turn into a project. TypeCal runs on a single repeatable rhythm.
Say it however it comes out. “two eggs and toast” is enough. Or point the camera, scan a barcode, paste a link.
Calories and macros arrive ready for review. Fix a portion, swap an item, then approve. Nothing counts until you say so.
Saved meals, recipes, and history make the next similar meal a two-tap job. The log gets faster every week.
Four ways in
Fast logging cannot force every meal through one rigid form. Pick the door that matches the moment.
Natural language for the meals you can describe faster than you could ever search a database.
Packaged food without the hunt.
The difference
Most AI trackers hide the guesswork. TypeCal shows its math — and when something looks off, you don’t re-log or argue with a chatbot. Tap the number, fix it, move on.
Every item, portion, and macro is editable before it saves. Numbers you’ve touched are numbers you trust — that’s what keeps a log alive past week two.
Join the waitlistProtein 8g Carbs 44g Fat 12g
320calories
The match, portion, and nutrition source at a glance.
The food match, serving, and nutrition profile align.
Why this result
TypeCal recognized Masala Toast / Sev Toast as Masala Toast with Sev. It scaled the matched nutrition data to the logged portion before totaling the macros.
Straight from the app
Start with the new TypeCal motion reel, then move through the interface running on iPhone today. Swipe through.
Food, words, and the current TypeCal interface — one continuous loop.
Calories, macros, water, weight — the charts fill themselves.
Your usual breakfast is a saved meal, not another search.
Reels and recipe pages become meals with per-serving macros.
Point it at a plate or a label. Same numbers, same review, same control.
Beyond the log
The details that make TypeCal useful long after the estimate.
01 / Kitchen
Import a recipe once, move what you need to the grocery list, and log a serving when it reaches the table.
Produce
Add grocery item
02 / Momentum
Food Story and streaks stay separate, just as they do in TypeCal. One remembers the day; the other marks the habit.
5 items logged
03 / Progress
Weekly calories, macros, and water use the same labels, colors, and target lines as the live progress screen.
1,842 cal
04 / Routine
Set a real meal window and use the same compact date control that lives beside the TypeCal composer.
TypeCal schedules one on-device reminder at the end of each enabled meal window.
05 / Essentials
Health, appearance, voice, and offline recovery now mirror the actual TypeCal states instead of decorative stand-ins.
You can still log your food. Your entries will resume syncing when you’re back online.
FAQ
TypeCal is launching on iPhone soon. The waitlist gets access the moment it goes live — we email you on launch day, nothing before that except the occasional note that matters.
There’s a free trial, so you can prove TypeCal fits your routine before paying anything. Beyond that it’s a paid subscription — we’d love to make it free, but every meal you log runs real AI, and that costs us real money. Charging honestly for it is how the app stays fast and keeps improving (it’s also why we can’t swing student discounts just yet). The price lands around one fancy Starbucks drink a month — the waitlist hears the exact number first.
iPhone first. Other platforms come later if TypeCal earns it.
Accurate enough that we publish the numbers. On a leading AI food journal’s own 68-meal benchmark, TypeCal’s estimates averaged 7% off the reference values — theirs averaged 11%. Every meal, hit or miss, is in the comparison tables. And you see every estimate before it counts, so any number is one tap from fixed.
Photo-first apps guess from a picture — and a picture cannot see the oil in the pan, the butter on the toastthe ghee in the dal, the oil in the pan, or how much you actually ate. TypeCal starts from your words, the most honest description of a meal there is, and shows you every assumption before anything counts. No database spelunking, no hidden pricing, and the accuracy claims are benchmarked in public.
Early access
Waitlist members are first through the door at launch, with the progress-first version of TypeCal we are shaping now.
Get early access