What's new 📣

The emoji you choose for your baby came in eight options, and exactly one of them was a baby. Unicode’s yellow baby is deliberately not a skin tone, so if yours did not look like the cartoon, your choices were a fox, a turtle or a teddy bear. There are now six babies, one for every skin tone, and thirty-two other options including a lot more animals: panda, koala, penguin, hedgehog, elephant, giraffe, octopus, butterfly, dinosaur and more. Every emoji that was there before is still there, so nothing you already picked has gone anywhere. The picker is the same on the add-a-baby screen and on Manage baby, which it quietly was not before.

Why: We built a food library that covers cassava, taro, callaloo and congee precisely so families would not have to abandon the foods they actually eat. Offering one baby, in one color, was a strange thing to have next to that. Thank you to the parent who pointed it out.

An expired or already-used confirmation link used to leave you on a spinning strawberry forever. It now says the link has expired, that nothing is wrong with your account, and gives you a way to get a fresh one. Changing your email now tells you the truth: we send a link to the new address AND to the old one, and both have to be clicked, which is what stops somebody else quietly moving your account to their address. If you have lost the old inbox, it points you at us instead of letting you wait for something that will never happen. Confirming a change now lands you back in Settings with a line saying it worked, rather than dropping you on Today with no word either way. Guests no longer see a change-password or change-email form, because a guest account has neither, and using that form could have lost the log. And a tab left open all day now notices within half an hour when your plan changes, instead of showing you Plus you no longer have or a paywall you have already paid past.

Why: Every one of these is a moment where you did exactly what we asked and the app went quiet on you. Silence after a correct action is the fastest way to lose somebody’s trust, and the hardest thing to notice from the inside.

The Redeem button in a gift email now carries your code to the page and fills it in for you, instead of dropping you on a page selling you the plan you were just given. If you are on a gifted year, the Plus page says so and tells you when it ends, rather than offering a Manage subscription button that answered “No subscription found.” Redeeming from a guest account is now refused with an explanation, because a guest account is cleared out when it goes quiet and the gift would have gone with it. If you are already subscribed, we will not open a second checkout and charge you twice. And when checkout itself is having a bad day, you get a sentence you can act on instead of a line of Stripe’s internal error text.

Why: Somebody spent real money on that gift. The least we can do is make the code take one tap, be impossible to lose, and never look like a sales pitch.

Four things about sharing, all in the same direction. If your Plus ends, the family feed page used to disappear entirely, while the people you invited carried on seeing every photo in it. Now sharing a new moment is the part that pauses; the list of who can see it, the Remove button, and taking a photo back down stay exactly where they were. The same on the care summary link for a sitter or pediatrician: you can still see and revoke a live link without Plus. Your viewer list shows the email address each person was invited at, because two people who never set a name both read as “Family” and Remove is permanent. Invites you sent that nobody has accepted yet are now listed, with a Cancel button, so an invite typed to the wrong address can be taken back instead of waited out. And a care link now expires when the person who made it loses access, so a sitter cannot leave a month-long link behind them.

Why: Paying us buys the sharing. It does not buy the stopping. Anything that hands out access to your baby has to be something you can take back on any plan, at any time, and you have to be able to tell who you are taking it back from.

The read-only summary you share with a sitter was built from a narrower slice of your log than it looked. It only counted bites matched to a food in our library, so anything logged by voice, offline, or against a food you added yourself was invisible to it. It also only counted a bite if the “first try” box happened to be ticked, so a baby with six logged peanut meals and the box never ticked showed nothing at all. Both are fixed: it now resolves foods the way the rest of the app does, by what the food is, and it reports the earliest date on file. When there is genuinely nothing recorded, it now says that is what the silence means rather than leaving the reader to guess.

Why: Somebody reads that page before they feed your child. An allergen quietly missing from it is the worst thing this app could do, and a blank space there must never be mistaken for “nothing to worry about.”

Our Terms have always said that deleting your account cancels future renewals when we bill you directly. The code did not do it. Deleting removed your account and everything in it, and the record of which subscription belonged to you went with it, so the card kept being charged for an account that no longer existed. It cancels now, on every route out, and it is written down the moment your account goes rather than depending on anything in your browser finishing. The delete screen tells you what will happen first: your subscription is cancelled, you are not charged again, and the part of the period you already paid for is not refunded automatically, so you can email us first if you want one. Same change on the other half of deleting: the photos and videos on a baby that goes with your account are now queued for removal by the server. They used to be cleaned up by your browser on the way out, which meant a dropped connection at the wrong moment could leave them behind.

Why: A subscription that outlives the account it paid for is money taken for nothing, and it is the kind of thing you find out about on a bank statement. A promise in the Terms has to be a thing the software actually does.

If your export names your baby and carries a birthday (Nara Baby’s does), the account step is filled in for you and the birthday is set, so nothing asks for it later. The page says when it is reading your file, the counts read properly (1 recorded first, not 1 recorded firsts), switching the solids option re-reads the file you already chose instead of blanking the page, and the same file can be picked again after Cancel. A file with nothing we can read now says so, and what to try. After the import, the card says your tracking history is in the log and, on the free plan, how many recorded firsts stayed in the file for Plus. One more, and an apology: the “confirm it’s you” link sent to a saved guest account opened as a page of code with no button, so nobody could confirm. It opens a proper page now. If you got the page of code, tap the link again, or use Send it again under Manage baby.

Why: A parent moving months of their baby’s life between apps should have to do as little as possible and be told the truth about what moved. Reading the name and birthday off the file removes two questions, and the success card should never say “everything” to a family whose recorded firsts are waiting for Plus.

When you save a guest account, we now send a short "confirm it’s you" email. Everything works straight away, logging, importing, all of it. The one thing that waits for that tap is inviting somebody else onto your baby’s log. If you need the link again, try an invite and we send a fresh one.

Why: An invite gives another person access to your baby’s history. Before anyone can hand that out, we want to be sure the email on the account really belongs to the person using it. Accounts made the usual way already work like this; guest accounts now do too.

Bringing your history over from another tracker used to mean making an account first and hoping it worked. Now you can drop your export straight in and see exactly what we found: how many feeds, sleeps, diapers and solids, and how many of your foods we recognized. Your file is read on your own device and never leaves it. If you decide to keep it, the account step happens right there on the same page, so the file you just uploaded is the file that gets imported. Your baby is called what you typed, not "Baby".

Why: Asking somebody to sign up before we have shown them anything is asking for trust we have not earned yet. You should be able to see that we read your file correctly, and that months of your history really are all there, while it still costs you nothing to walk away.

Two ways the same spoken food could come out wrong. Saying "Chinese five spice" logged it and saying "Chinese 5-spice" logged nothing, purely because the transcriber picked a digit that time. And "apple sauce" quietly logged Apple, a different food, because the library spells it "Applesauce". Now a digit and its number word count as the same word everywhere, including when you type in the food search, and apple sauce is apple sauce whichever way it is written.

Why: You said the food correctly. Which way our transcriber happened to spell it is not something you should have to think about, and it should never be the reason the wrong food ends up in your baby’s history.

The BLW-or-purees prep toggle now covers every single food we have a page for, 350 more of them today, from the everyday (brown rice, peanut butter, pizza, mac and cheese, grilled cheese) to the ones nobody else bothers with (kelewele, mujadara, attieke, chochoyotes, pounded yam, sixteen kinds of winter squash). Foods that should wait say so at every stage, drinks and spreads explain the form that actually works instead of pretending to be finger food, and every word is split from that food’s own already-verified prep and safety copy.

Why: Whichever way you are feeding, the answer to "how do I serve this today" should be on the page, for your food, not just the popular ones.

Every remaining thin page in the library is now a full guide: oca, ulluco, cloudberry, sea-buckthorn, pawpaw, ramps, alligator, and eleven more of the rare and regional foods. Each got a real intro, safety section, nutrition, pairings, storage, and six-plus FAQs, fact-checked against reputable sources before shipping. The two with genuine hazards say so plainly: ramps carry a sourced warning about poisonous wild lookalikes, and pawpaw notes the reason to keep it occasional.

Why: A family feeding their baby cloudberries deserves the same depth as one feeding bananas. This closes a campaign that started in July: zero thin pages remain.

The BLW-or-purees prep toggle now covers 70 more foods, starting with the ones families serve most: grapes (lengthwise quartering at every age, spelled out), peanut (the dissolvable-puff and thinned-butter forms by stage), plus farro, sausage, bacon, ice cream, sushi, and dozens more. Foods that should wait say so plainly at every stage instead of hiding it in fine print. Every entry is split from that food’s already-verified prep and safety copy, so nothing new to trust, just easier to use.

Why: The foods parents are most nervous about deserved the clearest by-age prep, not the barest.

New in Learn: "Baby blues or postpartum depression?" It covers the two-week line between the blues and something that needs treatment, the other shapes it takes (postpartum anxiety, the rare emergency that is postpartum psychosis, and the fact that partners get this too), why your baby’s checkups ask about your mood on purpose, and why treatment genuinely works. The helplines at the top are verified and worth saving before you ever need them. Every claim is sourced to NIMH, CDC, ACOG, AAP, and Postpartum Support International.

Why: You track your baby’s feeds, sleep, and moods all day. Someone should look out for yours.

Forty-four foods had prep guidance for 6 months and 12 months with a gap in the middle, right where the pincer grasp shows up and textures change fastest. Every one of them now has its 9-month step, in both the baby-led weaning and puree tracks, each derived from that food’s own verified prep and safety copy.

Why: Nine months is when "how do I serve this now?" changes its answer, and the ladder should not skip that rung.

On the Food tab, each food your baby has tried carries a "last had 12 days ago" line, and the Tried filter now sorts the longest-away foods to the top. So bringing a food back into the rotation stops being a memory test: open Tried, and the top of the list is exactly what has quietly dropped out of the meals.

Why: A food tried once and never served again is how a texture win or an allergen tolerance quietly slips. The record already knew; now it tells you.

The multigrain-cereal problem: logging it right meant picking spelt, barley, oat and rice one by one, every single morning, so most people logged it as one word and the allergen tracker saw nothing. Now you build the meal once, tap "Save as a usual", and name it whatever the box says. Next time it is one tap on the Solids tab, and every food still logs individually underneath, so a first taste of spelt still counts as wheat, firsts still fire, and every caretaker of your baby sees the same list.

Why: The right way to log a boxed cereal should not be the miserable way, and the allergen tracker should never miss a wheat exposure because it was 6am.

Logging "sesame oil" used to dead-end at "no match". The library now covers the oils real kitchens cook with: sesame (including toasted), coconut, avocado, canola, sunflower, vegetable, peanut, walnut, flaxseed, and mustard oil. The allergen handling follows the protein: sesame, peanut, and walnut oil count for allergen watches, because their common unrefined forms keep real allergen protein, and each note explains why oil is still a poor way to introduce an allergen on purpose. Plain vegetable oil does not count as soy, because highly refined soybean oil is fine for most soy-allergic people.

Why: If it goes in the pan, you should be able to log it, and the allergen tracking should follow the actual protein, not the label.

Tapping the "hasn't had fish in a while" notification dropped you at the top of the allergen radar, with the actual "easy ways to work it back in" list folded away further down the page. The tap now opens straight to that allergen's ideas, scrolled into view. And if you logged the food between the notification and the tap, the page simply shows your radar as usual.

Why: A notification that promises ideas should not make you hunt for them.

The optional details on a wet diaper offered the same fields as a dirty one, but size, consistency and mucus describe stool, not urine. Pick Wet now and those fields step aside; color and blood stay, because both are still worth capturing for a wet diaper. This also fixes a quieter problem underneath: the color notes read everything as stool, so a wet diaper marked pale could raise the pale-stool warning even though pale urine is completely normal.

Why: A tracker should only ask questions the diaper can answer.

A logged meal on the Today timeline listed its foods as plain text, so a thirteen-ingredient breakfast read as a wall of words while the Solids tab showed the same meal with every food wearing its emoji. The timeline now renders each food the same way the Solids tab always has, with the emoji glued to its name so lines wrap between foods rather than mid-pair.

Why: Emoji are how you scan a food list at a glance, and the home timeline is where you glance.

If you logged a meal after the fact, or edited one from earlier in the day, it jumped to the top of your recent bites as though it were the newest thing your baby ate. Breakfast could sit above dinner. The cause: the list was sorted by when you wrote the entry rather than when the food was eaten, and editing a meal rewrites it, which is why editing an early meal always threw it to the top. Solids now sort by the time you recorded for the meal itself, on the day it happened, so the timeline reads in real order. Feeds, sleep, diapers and meds were never affected.

Why: A feeding timeline that reorders itself when you correct it is worse than one you never corrected.

Every read on the Patterns page has an evidence floor, and the ones below it used to just name what they were waiting for. Now they also show your progress toward it: "already banked: 9 of the 12 evening feeds it wants." The counts only ever count up, deposits you have already made, and a card with nothing banked yet stays a simple sentence instead of a zero. No streaks, no guilt, no homework. Skipped a week? The week still happened, and the log picks up where you left off.

Why: These reads get sharper with every log you keep, and you deserve to see that adding up rather than take it on faith.

Seven new reads on the Patterns page (Plus), all from the log you already keep. Which bedtime slot has earned the longest nights in your house. Whether a feed close to bedtime has traveled with longer sleep, in your data, phrased as a pattern and never a cause. Which meal has been friendliest to new foods. When the biggest bottles cluster. Which foods graduated from a no to a yes this month, and which once-regular foods quietly fell out of rotation. And the careful one: if a fiber-friendly food like prunes or pears has been reliably followed by a dirty diaper in your log, Patterns will say it is worth noticing, with the receipts, and with real worries routed to your pediatrician where they belong. Every read stays silent below its evidence floor: a pattern from six data points is a coincidence with confidence.

Why: Every baby book tells you about babies. Your log should tell you about yours.

Two upgrades to the printable sitter page. Emergency numbers (911 and Poison Control, 1-800-222-1222, free and answered 24/7 in the US) now sit right on the safety card for everyone, free plan included. And family allergies got smarter: a food a parent is allergic to used to sit under "Do not give" forever, even after your baby had eaten it many times without a reaction. Now, once your own log shows the baby tolerates it, the sitter page moves it to a clear "already fine for this baby" note instead, keeping the family context without telling the sitter something your own records disprove.

Why: A sitter page has one job: being right in the moment someone else is holding your baby.

Reopening the app sometimes showed the add-a-baby welcome screen instead of your family, with tracker cards stuck on Loading until you force-closed and reopened. The cause: when the app resumed before your phone finished waking its connection, a failed data request was mistaken for "this account has no babies." Now a failed request is treated as what it is, the app retries on its own, and it remembers your family between launches so the real screen paints immediately.

Why: An app you open one-handed at 3am does not get to greet you like a stranger.

Two dozen foods in our library (raw fish, unpasteurized juice, hard candy, high-mercury fish, and friends) are ones we tell parents not to serve at all. You can still log them, on purpose: if someone slipped the baby a taste before you could object, that dated record is exactly what you and your pediatrician want. But the logger now warns you loudly at the moment you pick one, prepare-with-care foods show their safety rule right there too, and the quick try-it-out page no longer offers a cheerful first-bite form for foods that should never have a first bite.

Why: A tracker should record what happened without ever nudging you toward something unsafe. Recording is not endorsing.

Saying "Chinese five spice" out loud got you three puzzled suggestion chips instead of the food, because our food list spells it "five-spice" and the voice matcher took that hyphen literally. Speech never transcribes hyphens, so fourteen foods, including black-eyed peas, hard-boiled egg, and mahi-mahi, were unloggable by voice. Now a hyphen and a space are the same thing to the matcher, and all of them log on the first try. Caught by a parent logging a very ambitious dinner.

Why: Voice logging exists for the moment your hands are full of baby and dinner. It should not lose to punctuation.

Our diaper explanation used to say "iron-rich or dark foods like blueberries" can darken stool. Half right, fully imprecise: blueberries are actually low in iron. Their deep color comes from anthocyanin pigments, and those pigments are what pass through. Iron really does darken stool too, but from supplements and fortified cereal or formula, which is a different cause. The explanation now names the right mechanism for the food you actually logged. Caught by a parent using the app, which is exactly how this is supposed to work.

Why: A black diaper is a check-for-blood moment. The reassurance that talks you down has to get the reason exactly right.

Two things. First, the honest one: health photos were quietly failing to save, and we fixed it the day we caught it. Second, the good one: a new "Show the doctor" button walks through your saved photos oldest to newest, full screen with dates and notes, built for the exam-room moment. And a detail worth knowing: a photo taken in the app never lands in your phone’s camera roll (we checked, on iPhone and Samsung), but it does sync to your co-parent, so whoever takes baby to the appointment has it.

Why: A rash photo does not belong between birthday pictures, and describing a diaper from memory is nobody’s best moment.

Bites could already be saved offline and synced later. Now every log can: feeds, diapers, sleep (starting, stopping, or adding one after the fact), and medicines, whether you tap them or say them out loud. Anything saved without a connection shows on Today as waiting to sync, sends itself the moment you are back online, and never double-logs, even if the connection drops mid-sync. Undo works on queued entries too.

Why: The nursery at 3 AM is exactly where the wifi is worst and the logging matters most.

Two new guides show how feeds, naps, and solids commonly fit together at a specific age, each with a realistic sample day, the common wrinkle of that age, and what changes next month. They are examples, never assignments: no health body prescribes a nap count, and babies renegotiate daily. More ages follow if these earn their keep.

Why: Every schedule page on the internet is either feeding-only or sleep-only. A baby has one day, not two.

The 6-month-old schedule →

Two new pages anyone can read without an account: this changelog, at yummyyuc.ky/changelog, and a plain-words writeup of how we keep the food advice honest, covering the checks that rerun before every update, the every-three-months human review we cannot skip, and the corrections we publish. The dates at the bottom of every guide now come straight from that review clock instead of being typed by hand, so they can never quietly go stale.

Why: Trust us is not an argument. Here is the system, go check it.

How we double-check our food advice →

Voice logging a whole sentence at once, "napped 1 to 2, drank 4 oz, then pooped" becoming three entries, used to be a Plus feature. It is now free for everyone. Hold the Log button, say your piece, review what it heard, done. Transcription still runs entirely on your phone.

Why: The easier logging is, the more of your baby’s story gets kept. That should never have a toll booth on it.

Today has a new quick-log row: tap to log a wet or dirty diaper, start sleep, or repeat the last bottle exactly, each with an Undo right in the toast for the mis-taps that come with holding a baby. The log sheet now opens on the tab you actually use instead of always starting at Solids, so a pre-solids family lands straight on Feeds. Your usual bottle amount comes preselected. While sleep is running, the Log button shows a little sleeping face and opens straight to Sleep. And if you did not know you can hold the Log button and just say it, a one-time tip in the sheet now tells you. Android app-icon shortcuts (long-press the icon to jump straight to logging) arrive with the next app update.

Why: Logging a diaper was three taps. The things you do ten times a day should cost one.

Every baby app suddenly says AI, and the badge tells you almost nothing. The new guide walks through what these features really do, where modern AI genuinely helps (logging by talking, asking your own history), where the line sits between describing your data and diagnosing your baby, and six questions that beat the label. It ends with our own sheet filled out, because we make a baby tracker too and you should get to grade us with the same questions.

Why: Sophistication is not evidence that an answer is trustworthy, and parents deserve the difference in plain words.

Read the guide →

A lot of our food guides end with the same advice: skip the salty store version and make a simple one at home. Fair, except searching for the recipe found nothing. Ten of the most-asked are now in Recipes, free like all of them: baby tater tots, oven chicken nuggets, baked fish sticks, soft oven fries, little meatballs, first mac and cheese, oat and banana pancakes, no-sugar applesauce, frozen yogurt melts, and first guacamole. Each one handles the choking prep in the steps and names its allergens up front.

Why: A guide that says "make your own" owes you the recipe. One of us went looking for the tater tots one and it was not there.

Browse the recipes →

Pressing and holding a button used to make your phone try to select the label and offer cut and copy, which is never what a person holding a baby meant. Buttons, chips, and checklist rows now ignore text selection. Your notes, logs, and the guides all stay copyable.

Why: One of us kept hitting it while logging one-handed. If we hit it daily, you hit it hourly.

On the development checklist, the two "play along" ideas for each area now sit directly under that area, right where you just ticked. They used to live in their own section at the bottom of the page, below the earlier checklists, where honestly nobody scrolled. Nothing about them changed except the address.

Why: We built something genuinely useful and then hid it under the fold. Found it during our own use, which is embarrassing and also exactly how it should get caught.

The immunization record has moved out of the Plus checkup report and onto the Health tab for everyone. Add shots as they happen, or let an import bring them over, and print a clean one-page copy to hand a sitter, a daycare, or a new pediatrician. It also fixes something that was quietly unfair: importing your history from another app already brought your shot records in, but on the free plan there was no page that showed them. Your own records were in here and you could not see them. Now you can. It is a record you keep, not a schedule: your pediatrician owns the timing.

Why: The record is yours. Paying us should buy insight, never your own data back.

New comparison pages for Baby Connect and Glow Baby, written the way ours always are: every claim about their product read on their own pages with the date we checked it, and their strengths stated plainly. Baby Connect earns daycares; Glow earns its free breadth. One verified difference worth knowing: Glow charges for data export, and Baby Connect’s price is not published on their own site. Ours is free to leave and public to see. The privacy guide also gained the five questions everyone actually asks, answered in one line each: no, we do not sell baby data; no, there are no ad trackers; no, your voice never leaves the phone; caregivers see only what you share; and deleting your account really deletes it.

Why: A comparison that admits where the other side wins is the only kind worth reading.

There are 112 free recipes here, each one choke-safe, no added salt, allergens named, and until today there was no page that listed them. They were reachable from a single link on one food page and nowhere else. yummyyuc.ky/r now lists every recipe, it is linked from the bottom of every page, and the Learn page gained an A-to-Z list of all 139 guides so the newest ones are no longer findable only by search engine luck.

Why: We spent months writing recipes and then hid them.

Browse the recipes →

Small, medium or large, on any dirty diaper, next to color and consistency. Voice picks it up too: say “small hard poop” and both words land, where before the app kept “hard” and quietly dropped “small”. Saying “a little blood” still records blood, not a small poop. Size shows in the day’s list alongside the rest.

Why: Small and hard together is the classic constipation picture, and half of it was being thrown away.

Search for courgette and the zucchini page never said the word, and the same went for aubergine, capsicum, beetroot, rockmelon, tenderstem broccoli, spring onion, prawns and porridge. The regional names lived in our search index but not on the pages themselves. Eleven pages now say “also known as…” right under the title, each name verified against a dictionary or growers’ source first, because on a food-safety site a wrong “also known as” could merge two different foods. A new check enforces exactly that: it refuses any alternate name that already belongs to a different food.

Why: A parent in London searching courgette should land on the right page and see their own word for it.

Sugar apple, sweetsop, custard apple, cherimoya and soursop are the same family, and several are the same fruit under different names. Three of them told you to remove every seed, because annona seeds contain a neurotoxin and only the soft ripe pulp should be served. Sugar apple and soursop said nothing at all, so which warning you got depended on which name you happened to search. All five carry it now. Mochi and soda were also marked never-serve without saying why, and two high-mercury fish on the avoid list could not be found by searching their own names.

Why: Which name you happen to search should not decide whether you are told about a neurotoxin.

Logging a meal by voice and remembering halfway through that there was also toast is normal, and the app already handled it: tap the voice button again, say the rest, and it joins the same meal instead of starting a new one. You can type foods in alongside it, and go round as many times as you need before you save. It skips anything you have already added, so saying a food twice will not log it twice. None of this was written down anywhere, which is a strange way to treat one of the more useful things in here, so the voice helper now says so.

Why: Nancy pointed out that this is the bit that helps a tired brain most, and we had never mentioned it.

Typing cherry brought up cherry tomatoes, Cornelian cherry and ground cherry, but not plain cherries, because the library stores that one as “Cherries” and the search only ever tried turning plurals into singulars, never the other way round. It now tries both, so it no longer matters whether you type carrot or carrots, tomato or tomatoes, cherry or cherries. Near misses are handled too: chery, bananna, avacado and brocoli all find what you meant. Those only appear when nothing matched properly, so a real search never gets padded with lookalikes. Ranking is fairer as well: typing milk used to put Milkfish above whole milk, because any name starting with those letters beat a name containing the actual word. The word wins now.

Why: You should not have to guess whether we filed a food under one cherry or a bowl of them.

Attach a video to a bite and it saved and played back fine, but the recent bites list beside it stayed blank, so the meal looked like you had never added anything. The list was only ever looking for a photo. It now shows a tappable frame from the video, and tapping it opens the clip full screen the same way a photo does.

Why: You went to the trouble of filming it. It should be there when you look.

Every food started showing the same plate icon instead of its own, typing a food name came back “No match” for foods that are plainly in the library, and logging a bite by voice would hear you perfectly and then say it did not catch it. All three were the same fault: the app had quietly lost its food list. Two things caused it. The library passed a thousand foods, and only the first thousand were being sent, so the last twenty stopped existing anywhere in the app. And if the list ever came back empty for a moment, the app saved that emptiness and kept using it, so a single bad moment turned into a permanently broken food list on that device. It now loads the whole library however big it gets, treats an empty list as the failure it is, and repairs itself the next time you open the app. Nothing you logged was affected.

Why: Search, voice logging and every food icon all read from that one list, so when it went missing they failed in three ways that looked unrelated.

Your longest stretch against your best, wake-ups a night, wake windows next to the typical range for their age, and a note when the numbers look like a nap transition rather than one rough week. All of that has been free on every plan and will stay free. The card now says as much, and points at where the same treatment is applied to feeds, diapers and first foods, on What your logs say, which is part of Plus.

Why: You cannot tell from a features grid whether anyone’s analysis is any good. Sleep is the thing people most want read back to them, so that is the part we give away and let you check for yourself.

Journey had a card headed Milestones that awarded badges for logging: first bite, ten foods, a three-day streak. Milestone moments is where you record the developmental ones, first smile and rolled over and first steps. And the development checklist is the CDC’s list for their age. Tapping Milestones hoping for first steps and being congratulated on logging ten foods is a small letdown, and we were causing it ourselves. Journey’s card is now Badges, which is what it always was.

Why: Parents use that word for first steps. We were using it for a logging streak.

Both pages answer "how many a day?" from the same logs, and they were dividing by different things. The checkup report divided each figure by the days that figure was actually logged. Trends divided everything by one span, from your first logged day to today. If you log feeds daily and diapers a few times a week, that is not a rounding difference: nine changes across three logged days in a fortnight read as 1.3 a day on Trends and 3 a day on the report, and the report is the one you hand a pediatrician. Both pages now use the same rule, which is the report’s: each figure is averaged over the days it was logged. Both now say so on the page instead of implying the whole window, and Trends tells you how many of the last 7 or 14 days you logged on.

Why: Two screens showing the same baby should not disagree, and if one of them is going to the doctor it needs to be the honest one.

There were two printable pages for the same job and each was missing half of what the other had. The caretaker card listed the foods to avoid, the active allergen watch and the day’s notes, but not your pediatrician’s number. The care sheet had the phone numbers and the day’s totals, but its own thinner copy of the avoid list. Whoever you handed one to was short something. They are now a single sitter page: who to call, how the day has gone so far, what not to give, what we are watching, and today’s notes. The safety half is free, as it always was. Plus is the part that remembers your doctor, your emergency contact and your own instructions and prints them alongside. Handoff stays where it is, as the live screen you check on your phone.

Why: Nobody should need you on the phone explaining which of the two pages has the phone numbers on it.

Forty-one foods had a salt warning written on their guide page and nothing at all on the food itself, which is the bit that shows up when we suggest something, when you search, or when you log it. So the app could offer refried beans with no mention that the canned ones are often very salty, even though its own page says exactly that. Ham, bacon, kimchi, miso, sauerkraut, sardines, baked beans, hummus, anchovies, instant ramen and thirty more now carry the warning where you will actually see it. One entry was worse than silent: mujadara told you to serve a baby’s portion lightly salted, which is against both the NHS line and the food’s own guide. That is corrected.

Why: Babies should not have much salt because of their kidneys, and canned or cured versions of ordinary foods are where it hides. A warning only counts if it is attached to the thing you are looking at.

Every read on What your logs say answered the same kind of question: what has CHANGED. That needs two stretches of history to compare, which is why the page had almost nothing to say for the first six weeks, however carefully you logged. There is now one that answers what is TRUE instead, and it only needs about a week: the shape of a day right now. Awake around this time, this many naps, down around then, this many feeds and changes. It is the thing that is genuinely hard to see for yourself, because it is spread across a few hundred entries, and it is what most people want to know in the first month. It is a description of the stretch you just logged, never a schedule to hit.

Why: The page was built entirely out of reads that pay off at month three, which is a strange thing to show someone in week two.

What your logs say used to have one problem: most of it said nothing. A read that had enough history but found nothing moving looked exactly like a read that did not have enough history yet, because both came back empty, so most cards showed the same shrug about how either the history is young or there is no pattern. That was hiding real answers. If bottle volumes have held steady around 14 oz a day for two months, that is worth knowing, and for a lot of parents it is the reassuring thing they were quietly wondering about. Those now say so, with the numbers: diaper pace, night and day sleep, feed spacing, the milk and solids mix, bottle volumes. Reads that genuinely cannot speak yet are down to one line telling you what they still need, instead of a card each.

Why: Most weeks with a baby, steady is exactly the answer you were hoping for, and the page was throwing it away.

"When did they last eat" is the question this page exists to answer, and it had ended up eight cards down. It is now the first thing under any alerts: fed, awake or asleep, changed, each with how long ago, and each one a tap to log that thing again. The day itself moved up with it, so what happened today is near the top instead of past everything else. What moved down is the app talking rather than the day reporting: the tip, the resurfaced memory, the iron read, the suggested bite. All of it is still there, still goes quiet when it has nothing to say, and now reads after the day rather than in front of it. There is also a way back to the deeper reads from the bottom of the day: Patterns is now called What your logs say, which is what it does, and it ends with a line pointing at the rest.

Why: Every card on the page was worth having. The order had stopped working once there were eight of them above the thing you opened the app for.

What’s next used to show our own shortened rewrite of the CDC milestones: eight ages, paraphrased, nothing to tick. It now carries the real thing. All 159 milestones from the CDC’s Learn the Signs. Act Early. checklists, 2 months through 5 years, in the CDC’s words rather than ours, grouped the way the CDC groups them. Tick what you have seen and it stays ticked, so the list is ready when you are sitting in the waiting room being asked exactly these questions. You see the age they are in and the one coming next, not all twelve lists at once, and the earlier ages sit one tap away for the ticks that happen late. The four ages where the AAP recommends a real developmental screening are marked, so you know to ask about it at that visit. There is no score. We count what you ticked and stop there: no percentage, no verdict, no opinion from us about what an empty box means. That conversation belongs with your pediatrician, and the CDC’s own words about when to have it are on the page. All of it is free. The play ideas that go with each age are still Plus. When your child reaches a new checklist age we will say so once, and then leave you alone about it.

Why: Catching a delay early is the whole purpose of the checklist, and the earlier it is caught the more the help helps. That is not something to put behind a paywall.

We were calling a food a favorite the first time your baby liked it, which is how a cucumber tried once at breakfast ended up described as one of their favorites by lunch. Liking something once is not the same as loving it. A food now has to be genuinely taken to before we call it a favorite: enjoyed more than once, still enjoyed most recently, and enjoyed more often than it is refused. The bar also grows with your child, which matters more than it sounds. For a six month old just starting out, twice is a real signal. For a toddler with a thousand meals behind them, twice is barely a rounding error, so the bar rises quietly alongside their own history rather than staying frozen at a number we picked. Suggestions that pair something new with an old favorite now use the real ones.

Why: Getting this word wrong makes the app sound like it is not paying attention. Your baby has actual favorites, and they should be the ones we name.

If you are coming from another app, you no longer start from an empty screen. Export your data from Nara Baby or Huckleberry, upload it under More, and we bring your feeds, sleep, and diapers over into your log. You see a preview first, exactly what we found, before anything is saved, and re-importing the same file never doubles anything up. This is free: your history is yours, and moving it should not cost anything. Solids come over too, carefully. We match foods to our library only when the name matches exactly, never by guessing, and anything we do not recognize still comes across as plain history so nothing is lost. What we deliberately do NOT import: reactions, first tastes, or another app’s idea of what counts as an allergen being introduced. Those drive real safety surfaces here, and inheriting someone else’s version of them could leave you falsely confident that egg was already handled. Imported foods land as history, and the app asks you to re-check your allergen tracker yourself. It brings the whole day across, not just the daily loop: temperatures and medicines land in Health, pumping sessions with Feeds, growth measurements and shots in the record, and the color and texture of a dirty diaper now come over properly instead of being quietly dropped. Firsts your old app recorded come too if you are on Plus, since that is where the memory book lives; if you are not, we tell you how many are sitting in your file rather than losing them. Bath time and nail trims we leave behind, because we do not track those and there is nowhere honest to put them. The daily vitamin does come over, since that one we do keep, streak and all. From Huckleberry we now also keep how each sleep went: whether they went down upset, where they slept, and how they woke. In one real export that was 1,543 sleeps that used to arrive as bare times. And you no longer have to tell us which app you are leaving, or be leaving one we have heard of. Upload the CSV and we read the columns. One thing that quietly fixes: dates written day first. If your old app wrote 1/6 meaning the 1st of June, we used to read it as January 6th and file five months of your baby’s life in the wrong place.

Why: The hardest part of trying a new baby app is leaving months of logs behind. Now you do not have to.

Plus now belongs to the baby, not to one login. When anyone caring for your baby has Plus, everyone tied to that baby gets the Plus features for them: the co-parent, the grandparent you invited, the sitter. No more paying twice for one baby’s keepsakes, letters, and reports. Nothing changes about who can join, and only you can add someone (they come in through an invite you send), so your household is exactly the people you chose to bring in.

Why: We built the app around sharing the load. Charging each parent separately for the same baby was the opposite of that.

The home screen now offers one food idea a day, picked from your own log and explained in plain language: a new taste that pairs with a proven favorite, a flavor family they have barely met, a food that flopped once and deserves another go (most take 8 to 10 or more tries), or a common allergen not introduced yet, framed gently. The reason is the point: generic lists cannot know your baby, and this one does, because you built the record it reads. The pick holds steady all day and rotates tomorrow, one tap logs it, a “not this week?” tap snoozes a suggestion for seven days, and it stays quiet before solids or when there is nothing honest to say. When your log shows first tastes landing best at a certain time of day, a new idea arrives with that timing attached. The daily idea and its reason are free; Plus adds the runner-up ideas with their reasons. And for texture-ladder families: rung firsts (first purees, first lumpy food) now come back around as anniversaries with the other memories.

Why: The answer to "what should I try today?" was already sitting in your log. Now the app says it out loud.

Five new pattern cards read the arcs parents feel but never see measured. Milk to meals: bottles easing down while solids climb, the normal shift toward table food told from your own log. Night feeds: about one a night lately, from three two months ago, possibly the best sentence a tired parent can read. Feed spacing: daytime feeds stretching from 2h 40m apart to 3h 20m. Morning wake-up: settled within about 20 minutes of 6:40, because predictable is plannable. And the nap transition read now sits on the Patterns page too. On the Health page and the pediatrician report, a finished fever now reads as one story instead of scattered rows: when it started, how high it ran, what was given, and when readings came back to usual, in your °F or °C. The texture ladder starts keeping its own timeline (purees to lumps to finger foods, dated) from your next change. And the monthly letter arrives: each month of the log written back as a keepsake, with the first tastes, the longest night, the milestones, and your own words from the portion notes. Printable, and a quiet month simply writes no letter rather than a hollow one.

Why: You lived these months. The log was already keeping them; now it can hand them back.

The little "see the data" toggle from the Patterns page now lives under the insights you already had: the next-nap card shows what its estimate rests on (their own last wake windows, or age norms, and how many), and the keep-offering page shows the counts behind the timing read, the favorite-on-the-plate read, and your median tries to a win. The weekly digest email gains one line per baby when there is a real one to say, like "Sunday nights are usually Poppy’s longest stretch", with the numbers, because a recap you already lived through is nice but a pattern you had not noticed is a reason to open the email. The pediatrician report adds a two-month trend line: how the longest sleep stretch and the daily bottle total have moved, counts only, no judgments. And the favorite-on-the-plate read now speaks for moderate meal-loggers too, not just the most devoted: it needs five paired tastes now instead of eight, and the receipts show exactly how many it rests on.

Why: Same promise as the Patterns page, applied everywhere: a plain sentence up top, and the data underneath for whoever wants to check our math.

A new page for the spreadsheet parent in your life (it might be you). Patterns reads the logs you already keep and says things you could technically work out yourself but realistically never would: which night of the week usually runs longest, how bedtime has drifted over the last month, whether a shorter first wake window tends to mean a longer first nap for your baby, how your diaper pace is changing, how many tries a new food usually takes in your house, how bottle volumes have moved (bottles only; nursing is real food the math cannot see), how sleep is consolidating into the night, and, once you have thirty days of history, how deep the log itself has grown: months, entries, and foods nobody else has. Every card leads with a plain sentence, and a little "see the data" underneath shows the receipts: the medians, the windows, and exactly how many nights or mornings each claim rests on. Three promises hold everywhere. These describe your baby’s own history, never a comparison to other babies. A pattern is not a cause, and the cards say so. And a card with too little history, or no real pattern to report, stays quiet, because silence is more honest than a stretch. Nothing new to log; more history sharpens every card on its own.

Why: You gave the app the data. It owes you something interesting back.

If you started as a guest from a diaper or bottle page, the app used to greet you with the full solids experience: bite counters, first-foods checklists, allergen cards, none of it relevant to your two-week-old. Now, until we know your baby’s age, you get the quiet version: feeds, sleep, diapers, and nothing else. The getting-started list gains one new step, the only question we will ever put in front of you rather than work out ourselves: the birthday. One date, typed right there on the page, and nap timing, solids readiness, and age-right guidance all tune themselves to your baby. Skip it and everything still works; the food tools appear on their own the moment you log a first bite. And when a guest saves their account, there is finally a spot to give your baby their name, so nobody is stuck with a keepsake book about a child called Baby.

Why: A great first visit means showing the right things, not all the things. One optional date does more tuning than any questionnaire.

Feeds have always let you pick ml or oz, but weight and temperature quietly did not. Weight took pounds and ounces only, and the thermometer field took Fahrenheit only, so if you think in kilograms and Celsius you were doing conversions in your head at 3am, or you simply could not log a temperature at all: the safety check that rejects impossible readings was written in Fahrenheit, so a perfectly normal 37 got refused as implausible. Both now have a unit switch right next to the field, and the app starts on whichever system your device already uses. Nothing you have logged changes. Weights and temperatures were always stored the same way underneath, so switching just changes what you type and what you read, on that device, and your co-parent can read it the other way on theirs.

Why: Most of the world weighs babies in kilograms. Refusing their numbers was not a preference, it was a bug.

Plenty of babies react to milk, soy or egg long before solids start, through breast milk or formula, and until now the allergen radar had no idea. It would cheerfully put a gold star on the exact food your pediatrician told you to hold. You can now flag any of the nine as one your baby has reacted to, including how it showed up. We stop suggesting that one, drop it out of the rotation reminders so nothing contradicts itself, and carry it into the pediatrician summary with the date and the route. The important part is what we do NOT do: holding one allergen never holds the others. Guidance is clear that delaying the rest raises risk, so the radar keeps the other eight moving on their normal timing and says so on the page. We are not writing the plan here. Your clinician does that, and we keep the record.

Why: One bad early reaction makes it tempting to pause everything. That instinct is the one thing the evidence argues against, so the app should hold exactly what your doctor is holding and nothing more.

Iron is the nutrient that matters most in the first year: a baby’s stores run down around six months, and the CDC recommends iron-rich foods daily from then. So Today now shows a small card counting the days iron-rich foods appeared in your log, with examples, and the pediatrician Checkup report carries the same count over two weeks, sourced to USDA data food by food. One thing we will not do, on purpose: tell you whether your baby is getting “enough”. We do not know your portion sizes, so we cannot judge amounts, and we will not pretend to. What we can say truthfully is which days iron showed up, that meat, poultry and fish absorb best, and that beans, lentils and fortified cereals count, all per the CDC. Your pediatrician does the interpreting; we just bring the record.

Why: Iron worry is real and mostly invisible. A day count you can glance at, and a line your doctor can read, beats both silence and false precision.

The firsts collection only knew food, but some of the most-told stories in parenting were sitting in the sleep log: the first 4-hour stretch, the first 6, the first time through the night. Those now appear on your Firsts page, dated from your own history, and the moment a new longest-ever stretch is logged you get the same celebration a first taste gets. On its anniversary, the memory comes back around: "a year ago tonight, the first time through the night." A few more quiet additions in the same spirit: the little notes you write behind the portion chips ("didn’t want to stop eating applesauce") now resurface as memories, a food that grew from a single taste into a full serving gets its story told, and a month of daily vitamins with no gaps earns a word of recognition, since nobody else hands out medals for that. One promise across all of it: sleep records only ever speak on good news. A short night is just a night, and the app will never tell a tired parent their baby’s sleep is getting worse.

Why: You were already writing this diary. We just started reading the whole thing back.

Three quiet additions that all come from your own log. First, the app now tells you the moment it notices a pattern, instead of leaving it on a page you would have to think to visit: the first time it can see when new foods land best, or how many tries foods usually take for YOUR baby, or whether a favorite on the plate actually helps, a small note appears on Today, once. Second, that last one is now checked for real: because meal logging records which bites shared a plate, we can compare new foods served next to a loved food against new foods served without one, in your own history. And third, the memory resurfacing learned the best story it can tell: on the anniversary of a refusal of a food your baby now loves, it hands the whole arc back. "Broccoli got a firm no, twice. Now? A favorite." Everything here stays silent until there is enough history to be worth saying, because a pattern invented from six bites would be worse than none.

Why: Articles can tell you about babies in general. The only thing that can tell you about YOUR baby is their own record, and it should speak up when it has something, not wait to be asked.

Almost every bite you log carries a clock time, and until now nothing read it. Now, once there is enough history, Keep Trying shows when first tastes actually land for your baby: "27 of 30 new-food tries got a yes in the morning, against 3 of 7 in the evening." The logger uses it too, so after a rejected bite the encouragement can add the one useful lever you have besides patience: the time of day that has been working. It stays completely silent until at least two parts of the day have enough first-taste history to compare, because a pattern invented from six bites would be worse than no pattern.

Why: Every feeding guide says "keep offering." None of them can tell you that YOUR baby says yes at breakfast and no at dinner, because that answer only exists in your own log. This is the kind of thing we can know about your baby that no article ever could.

Tell us how many wipes you have and we will tell you when you are about to run out, using the diaper changes you already log. Nothing extra to track. It also counts properly: a dirty change burns several wipes and a wet one burns a couple, so the estimate follows whichever phase you are actually in rather than assuming an average. On real logs that difference is about 28%, which on a big pack is a few weeks either way. The nudge is deliberately gentler than the diaper one, since running out of wipes means reaching for a flannel, not an emergency.

Why: Wipes are the thing everyone buys constantly and nobody thinks about until the box is empty at 2am. You were already giving us the information needed to see it coming.

The What’s next page pairs the CDC milestone checklist with things to try at home. The checklist half was always complete; the doing half was two ideas per stage, and a stage lasts three months. It is now eight per stage, 64 in total, sorted under the same four headings the checklist uses so each idea is attached to the thing it helps with. At 9 months, peek-a-boo sits under Social because that is the month they work out you still exist when you leave the room, and rakeable finger foods sit under Movement because raking food is literally on the list. They are ideas, not a second checklist. Skip any that do not suit your week.

Why: Anyone can find a milestone list. Knowing what to DO about it is the part worth having, and two suggestions for a three-month stretch was not that.

The last 136 food guides that had never been fact-checked have now been read in full, not skimmed by a script. Most needed nothing, which is the result we wanted. Five did. Walleye was listed as a low-mercury everyday fish in one part of the app and a moderate-mercury occasional one in another; the cautious version was right, and the app now agrees with itself. Perch was described as a once-a-week fish when it is actually in the FDA lowest-mercury group you can serve twice a week. The morel page said raw morels cause an upset stomach, when the outbreak we now cite killed two people, so it says that instead. And rhubarb, swiss chard, veal, quelites and morogo were the only guides of 939 that never told you WHAT AGE to start, because their prep sections were written as steps instead of ages. They have proper age ladders now. Separately, 79 guides made a correct safety claim while linking to a source that did not mention it, so a parent clicking through to check us would have landed somewhere unhelpful. They all point somewhere that backs the claim now.

Why: You should be able to click any claim we make and land on the page that proves it. That only works if someone has actually read every page, so we did.

Our pasta guide already told you to reach for a big grippable shape like fusilli or penne, cooked well past al dente, because a six-month-old grabs with a whole fist rather than finger and thumb. What it did not cover was the chickpea and lentil pastas that are now on every shelf. They are worth knowing about: the texture cooks denser and stickier, which is harder work for a beginner, but the bigger point is that legume pasta is not really pasta, it is legumes. Chickpea, lentil and pea proteins cross-react with each other and with peanut, so it is a new allergen introduction on its own day, not a swap for the wheat pasta they already eat.

Why: Nancy went looking for pasta advice and found a decent answer that stopped short of the thing that actually matters. Swapping in a "healthier" pasta feels like a nutrition decision and is quietly an allergen one.

Until now the Flavor Explorer counted the foods you had tried in each flavor family and ignored how they went, so "explored bitter greens" could mean "met kale once and hated it". It now shows how many of the foods you offered in each family actually landed, as a plain count. If one family is running low and another is going well, it suggests bridge foods: things that sit in both, so you are offering half a flavor they already trust instead of another lone green. A low row is a place to keep offering, never a verdict on your kid.

Why: Where you have been is a map. What they like is a read on your actual baby, and it is built from months of your own logging that nobody else has.

Solids, sleep, feeds and diapers each answered questions about themselves and never talked to each other. The recap now compares your week against the three before it and names anything that moved a lot, side by side. Small wobbles are ignored on purpose, because baby data is noisy and a page that announces every 5% change is a page you learn to skip.

Why: The interesting weeks are the ones where two things moved together. We will show you both numbers and let you make of them what you like. We are deliberately not telling you one caused the other, because a feeding app cannot know that, and anything genuinely worrying belongs with your pediatrician.

The toddler meal ideas list had 17 entries, which is about a week before you have seen all of them. It is now 63, spread evenly across breakfast, lunch, dinner and snacks, and it still leads with the meals built from foods your toddler already eats. Every one names how to serve it safely, and anything round still tells you how to cut it.

Why: A toddler eats roughly a thousand meals a year. Seventeen ideas is a rounding error, and running out of suggestions is the whole reason people stop opening a meal planner.

When you log a 😐 or a 🤢 on a food that has not landed yet, the logger tells you how many times you have offered it, and what that number usually means. Research says a new food often takes 8 to 10 or more tries, and most of us give up around three. The Keep Trying page also learned to do the maths on your own baby: once five foods have been won over, it tells you the median number of tries YOUR baby needed, which beats any number from a study. Nothing appears for a food your baby liked straight away, which is most of them.

Why: The moment you decide to stop offering a food is the moment you are staring at a face full of rejected broccoli. That is exactly when it helps to know you are on try four of a normal eight, and that the parents who get there just kept going.

Searching the recipe list for an allergen only found recipes that happened to spell it out, which is almost none of them. "Wheat" returned 2 recipes when 16 contain wheat, because the other 14 say bread, pasta, toast or udon. "Sesame" found 1 of 4, since the rest say tahini or hummus. Soy and fish found none at all. Milk looked like it worked and was finding 9 of 20, because cheese, ricotta and butter never say milk. Search now understands what an ingredient actually contains, and each result tells you why it matched, so a hit on hummus shows "uses hummus". While we were in there we also stopped the search for "egg" from returning eight vegetable recipes, which it had been doing on the "egg" inside "veggie".

Why: Nancy noticed tahini and sesame were coming up short and asked whether wheat returning 2 was right. It was not: it should have been 16. If you are searching an allergen you are usually either trying to introduce it on purpose or trying to find where it is hiding, and both of those need the whole list.

In the allergen section, egg had three forms to work through, and the third was "Soft or runny" sitting there with a "not tried" badge like a box to tick, even though its own description said raw or barely set egg is not for babies. It now reads "Runny or barely set" and is badged "not for babies" instead, and nothing in the app will ever suggest it as your next step. We also fixed the sorting: a soft scramble that is cooked through was being filed as runny, when it is simply a well-cooked egg and counts as one.

Why: Nancy spotted it and put it plainly: we were giving people credit for trying something we had just told them not to do. A checklist that rewards the one thing on it you should skip is a checklist working against you.

We went through our highest-risk food guides line by line against the FDA, CDC, NHS and USDA, and fixed what did not hold up. Five pages listed the high-mercury fish to avoid and stopped short of the full list, which is seven species and includes bigeye tuna, the thing a lot of "fresh tuna steak" actually is. Grouper and snapper gained a note about ciguatera, a reef-fish toxin that cooking does not destroy. Mussels gained one about biotoxins, and got a real cooking endpoint instead of "until the shells open", which is where cooking starts, not where it ends. Rhubarb, elderberry, raw bamboo shoots, raw flour and deli meat all had serious warnings written on their pages but nothing attached to the food itself in the app, so the warning never travelled with the food. Abalone and octopus were calling themselves top-9 allergens, which molluscs are not. And a few pages disagreed with each other about ages, so cooked shellfish now says the same thing everywhere.

Why: We tell you a food is safe or not safe, and you act on it that day. That only works if we have actually checked, so we did, including the parts where we turned out to be wrong.

Real guides for the foods the app was happy to suggest but could not explain: gungo peas, string cheese, baby corn, snow peas, hake, blackcurrant, capon, frog legs, tamale, dolma, sambar, champurrado, risotto, gnocchi, pierogi, frittata, omelette, and the baking flours (wheat, whole wheat, oat, almond) plus cornmeal. Each one covers when to start, how to prep it by age, and the specific thing that makes it risky, like the fact that a string cheese stick is the classic choking shape and has to be pulled into strands, or that raw flour is a raw food and not a play dough.

Why: Suggesting a food and then telling you we have nothing to say about it is worse than not suggesting it. If it is in the app, it should come with instructions.

We had written full guides for all 35 spices and herbs, nutmeg included, and the app was not linking to any of them: tap a spice in your food library and it looked like no guide existed. Fixed. Variety foods now point at the right guide too, so "blueberries" opens the blueberry guide and "basmati rice" opens the rice one instead of a dead end. And cornmeal porridge finally has a recipe to go with its guide, because "cook it soft and unsweetened" is not much help if you have never made it.

Why: Nancy hit all three in one sitting: a spice with no guide, a suggested food that led nowhere, and a recommended breakfast she would have had to go and look up how to cook. Suggesting a food and then leaving you to figure it out is not a feature.

If you asked for a guide on a food that does not have one yet, the button told you to sign in even when you already were, and the request was never recorded. It is fixed: your request is counted now, and the sign-in note only appears when you are genuinely signed out. We also corrected our nutmeg advice, which said to skip it entirely on cornmeal porridge while suggesting a pinch of it elsewhere. A pinch is fine from about 6 months; it is a heavy hand that is worth avoiding.

Why: A button that silently swallows what you asked for is worse than no button, and advice that contradicts itself between two pages is not advice you can trust.

The search box on the Library tab now covers recipes alongside foods, guides, and spices. Type a recipe name or just an ingredient ("sweet potato") and matching recipes appear right in the results, including the ones you have saved yourself. The Recipes tab keeps its own search too.

Why: You should not have to remember which tab something lives on. One search box, everything in it.

Plus now builds you a week of meal ideas from our recipes, and it is smart about your baby. It keeps the allergens you have already introduced in rotation (most overdue first), works in the taste families they have not explored yet, and leans on the foods they love. Every meal tells you why it is there, so it is a plan you can trust rather than a random list. Shuffle it any time.

Why: Two of the hardest parts of feeding a baby are "what do I make this week" and "am I keeping their allergens in." This answers both at once, from your own baby's history, instead of a generic menu.

Log a few sleeps and Today shows the next likely nap window by day and a reasonable bedtime by evening, read from your own baby's wake-window rhythm. This used to be a Plus feature. It is now free for everyone. Wake-window trends in Trends are free too. It is a heads-up, never a rule.

Why: Guessing when the next nap lands is one of the most stressful parts of a baby's day, and other apps lock it behind a subscription. Ours reads your baby's real pattern, and there is no reason to paywall a calmer afternoon.

We corrected the allergen tags on oatmeal and barley groats: oats are gluten-free and not a top-9 allergen, and barley is a gluten grain but not the wheat allergen. Neither counts as a wheat introduction anymore. If your tracker now shows wheat as not-yet-introduced, that is why, and it is the safer read: to actually check wheat, offer a wheat food like bread or pasta.

Why: Logging oatmeal used to give false credit for introducing wheat, which could leave a family thinking wheat was safe without ever testing it. Accurate allergen data is not optional.

Baking with your little one? You can now log wheat flour (all-purpose or whole-wheat), oat flour, almond flour, and cornmeal. Wheat and almond flour count toward those allergen introductions, so a batch of banana pancakes shows up as the wheat exposure it actually is.

Why: Real feeding happens in real kitchens. If a food goes in the mix, you should be able to track it, especially the allergen ones.

We more than tripled the built-in recipes (from 17 to 60), spanning first purées (avocado-mango, pumpkin-apple, beet-apple), 9-month combos (soft egg and avocado toast fingers, salmon and sweet potato, dal and rice, chicken congee, plantain and black bean), and toddler dinners around the world (rice and beans, a mild jollof-inspired chicken and rice, a soft peanut noodle bowl, turkey meatballs, beef and bean chili). Each is choke-safe, no added salt, no honey, with allergen notes, and built from foods already in your library.

Why: The daily "what do I actually make?" deserves real, safe answers, not just single ingredients. Recipes turn the food list into meals you can put on the table.

When refusals pile up, Today gives you a gentle heads-up, and Plus adds a coaching page for riding out a food jag: the one rule that defuses mealtime battles (you decide what and when, they decide whether and how much), what actually helps, and the specific foods your little one is pushing back on right now, with anchor foods to serve alongside. Sourced, and honest about when it is worth calling your pediatrician.

Why: Food jags are normal and they pass, but in the moment they are stressful, and it is easy to slide into pressure, which backfires. A calm, evidence-based plan beats a mealtime battle.

A new Plus surface for the "what do I even feed them" stage: real, balanced toddler meals from breakfast to dinner, ranked so the ones your toddler already has the ingredients for come first, each with one gentle stretch food to keep trying. Curated by us, never a random combo generator.

Why: Toddlers eat like little people now, and the daily "what's for lunch" gets old fast. Starting from the foods they already like makes the answer easier, and sneaks in one new thing at a time.

Once your little one hits a year, Today starts speaking toddler: a gentle heads-up that days settle into about 3 meals and 2 snacks, that the sudden new-food wariness is normal and passes, and a nudge toward Keep Trying for the foods you're working back in. Same calm app, meeting you where your kid actually is.

Why: First foods were never the whole story. The toddler stretch is where feeding gets tricky in new ways, so the app should grow up alongside your kid instead of staying stuck on purées.

In the evening, the sleep card now shows a bedtime window instead of a nap time. It's anchored to when your baby usually wakes in the morning and the sleep they need at their age, so it stays put instead of drifting later every time a nap runs long. Not enough sleep logged yet? It falls back to a typical range for their age.

Why: A wake-window nap timer will happily suggest a 10 PM 'bedtime' after a late nap, which is backwards. Bedtime should anchor to the morning, so now it does. A guide, not a rule.

Explore has a new free Textures tab: the steps from smooth purées to mashed and lumpy, soft finger foods, mixed textures, and family food. Tap each one you've offered and it fills in, with a link to the sourced guide. Tracking each rung is free. Plus adds step-by-step coaching for getting from one texture to the next, and lets you keep a rung as a dated milestone. It sits right next to the allergen tracker, because texture is the other half of eating well.

Why: Texture quietly trips people up more than flavor does, and there is a sensitive window for lumps around 9 months. This is a calm, no-pressure way to see where you are and what is next. It is a ladder to climb, never a test.

Milestones now hold a photo or a short video, not just words. Add the shot of the first wobbly steps or the clip of the first laugh right on the milestone, dated and age-stamped. We also drew a clearer line between the two keepsakes: Firsts is for first foods, Milestones is for the big developmental moments, and each now points to the other so you always land in the right place.

Why: A parent went to add a picture to 'first steps' and there was no way to do it. That is exactly the moment a keepsake is for, so now the photo (or the video) lives on the milestone itself.

We added a big batch of world foods, so the library now covers more than 1,000. That includes regional first foods and dishes (uttapam, jollof rice, ugali, milkfish, makhana and more), each with prep by age, choking notes, and honest safety sourcing. Two new deep-dive guides came with it: Indian first foods and West African first foods.

Why: Babies eat what their family eats. The library should treat jollof and dal as normal first foods, not exotic curiosities, so more families see their own table in it.

When something looks off, a bloody or unusual diaper, or a rash, you can now save a private photo in the Health tab to show your pediatrician. It stays walled off from the rest of your logs and your keepsakes, visible only to your family, with an easy per-photo 'only me' option.

Why: A photo says more than a 2am description ever could, and you should be able to keep it somewhere private and dated instead of buried in your camera roll.

You can now get YummyYucky for iPhone from the App Store, alongside Android and the web. The same calm whole-day tracking, sourced food library, and allergen tracker, now a tap away on your home screen.

Why: iPhone was the most-asked-for place to have us, and some days you just want the app on your phone without thinking about it. Now it is there.

If you track your diaper or formula stash, Plus now sends a gentle heads-up a few days before you run low, with a tap to the best price. It reads how fast you actually go through them, so the timing is yours, not a guess.

Why: Running out always happens at the worst moment. A few days of lead time means you buy at the good price, on your schedule, instead of a panic order at 2am.

Invite grandparents and family to a private, read-only feed of the moments you choose to share: photos and little notes. They sign in with a tap, no password, and they only ever see what you share, never your trackers or health. Nothing touches social media.

Why: The people who love your baby want to watch them grow, but not everyone belongs on a public feed. This is a calm, private place just for them.

Your savings page now shows a restock radar: when your diapers or formula will likely run out, read from how fast you actually go through them, plus the best price to catch right now. Enough lead time to skip the emergency run.

Why: Running out is always a surprise at the worst moment. Now it is on your radar a few days early, when you can still shop the good price.

A gentle, age-by-age guide to what most babies are doing right now (from the CDC's milestones), plus a few easy ways to play along and a peek at the leap that's coming. Tuned to your baby's age.

Why: You wonder if what you're seeing is on track and what to try next. This gives you a warm heads-up instead of a 2am search spiral. It's a guide, never a test.

Record short clips of the first laugh, the babble, a bedtime song, right in the app, and keep them alongside your other keepsakes. Play them back any time.

Why: A photo freezes a face, but the sound of them at this age is the thing you forget first. Now you don't have to.

A little time capsule: write notes to your child to read years from now, tagged to a moment like a first birthday or their wedding day. They stay safe with you until you decide to share them.

Why: The things you most want to tell them are easy to think and hard to keep. Now they have a home that only gets more precious with time.

Hand off one clean page: today so far (feeds, naps, diapers), the doctor and an emergency contact, the foods to avoid, and your own instructions. Print it or save it as a PDF.

Why: Whoever has your baby shouldn't have to text you fifteen questions. Now the answers are on one page.

While your baby is awake, Today can now show when the next nap is likely, read from their own wake-window rhythm and their age. If they have been up longer than usual, it gently flags that too. A guide, not a rule.

Why: Catching the sleepy window before the overtired meltdown is half the battle, and this reads it off the naps you already log.

Once a week we send a gentle recap of how it went: the meals logged, the new foods tried, and a tap straight to the full picture. Someone should remember all of it. It doesn't have to be you.

Why: The app already quietly holds your whole week. It should hand it back to you now and then, instead of making you go dig for it.

Your checkup report now opens with a quick 'before your visit' summary: what changed since last time, and a short list of things worth asking about, all pulled from what you already logged. Conversation starters, never medical advice.

Why: The questions you meant to ask are the ones you forget once you are in the room. Now they are waiting at the top of the report.

Turn every first, milestone, and growth note into a designed keepsake book you can print, save as a PDF, frame, or gift. Find it from your weekly recap.

Why: The trackers fade, but the keepsake only gets more precious. It deserved to become something you can hold.

When more than one person is caring for your baby, the handoff screen now shows a warm recap of everything the other caretakers logged while you were out: the feeds, naps, diapers, and first bites, and who did what.

Why: Coming back after a few hours away and piecing together what you missed is its own little stress. Now it is just there, already added up.

The Flavor Explorer's world passport can now be saved as a designed card to share, with every region your baby has tasted lit up. A new badges shelf collects the milestones as you go (First Bite, Brave Eater, World Traveler, a full passport, and more), and the map gives you a little celebration the moment a new region lights up.

Why: Tasting the world is one of the happiest parts of starting solids. It should feel like filling a real passport, and be something you can actually show off.

Everyone can now save a designed keepsake card of their firsts to share, no more plain text. And Plus brings the firsts to life: a full-screen reel of every first-bite face, a little "food story" of the biggest hits and misses, a printable poster to frame, and a gentle "on this day" note when a first hits an anniversary.

Why: You told us the firsts are the thing you will actually treasure years from now. They deserved more than a scrolling list.

Set a reminder for anything on a schedule: the daily vitamin D, a twice-a-day nebulizer, or a 10-day course of antibiotics with the days counting down. You pick the med, the dose, and the times, and we send a notification when a dose is due (and flag it in Health) so you can log it in a tap. It is a helper, not a substitute for medical advice, so keep following your doctor and the label.

Why: "Did I already give it?" is one of the easiest things to lose track of when you are running on no sleep. Now the app holds the schedule so you do not have to.

Feeds has a new 'Same as last' button. Tap it and the form fills itself in from your previous feed, the same bottle or nursing, ready for you to log.

Why: A baby feeds 8 to 12 times a day, usually the exact same thing as last time. Re-typing it every round was busywork nobody needs at 3am.

Ending a nap or logging a sleep now shows an Undo, the same as the other trackers. Tap the wrong wake time and you can put the timer right back.

Why: You are usually logging sleep while short on it, so a one-tap take-back means a fat-fingered tap is not a chore to fix.

The second you log a food the baby has never had, a card pops up to grab a photo right there. Missed it in the moment? Open Firsts and tap the empty frame on any first to add a photo later, even one from your camera roll.

Why: Those first faces are the thing you will actually want years from now. We did not want them buried two taps deep in a details screen while you are wrangling a spoon.

A real meal is rarely one food. It is oatmeal plus banana plus a shake of cinnamon. Now you build the whole plate from one box, tapping recent foods instead of typing each one, then rate it all at once.

Why: Babies move to mixed meals fast, and the old flow made you log them one food at a time. Every ingredient still gets tracked on its own for the allergen tracker. You just do it in a fraction of the taps.

Enter the ounces of powder you have on hand and we tell you how many days it lasts, in ounces and milliliters, tuned to your brand's scoop yield.

Why: You buy formula by the can of powder, not by prepared ounces. The math to bridge the two is fiddly and easy to get wrong right when you are running low and least want a surprise.

Photos are now resized and compressed on your phone before they upload, so they post quicker and use far less data.

Why: A first-bite photo should attach in a blink, even on the crowded wifi in a pediatrician's waiting room.