Every Secondary Active Transport tool now carries a small pill on its home page: green "Signed in" when your answers are syncing to your account, amber "Not signed in" when they're saving to this browser only — and the amber pill is a one-click path to signing in. It stays out of the way while you're inside a quiz or an exam skin. No more finishing forty questions before discovering they were never going to follow you to your other device.
In every Secondary Active Transport QBank, reviewing your incorrect questions had a frustrating quirk: get one right this time and it jumped straight to the next question without showing the rationale. The review list was quietly dropping the question the moment you redeemed it. The list you're working through now holds still until you leave, so every answer shows its full explanation.
The mark beside Active Transport is now literally the transporter from the loading screen - the same drawing, scaled down. A substrate molecule is pumped up from below the membrane, through the pore, and released off the turning head as ATP: it goes in blue, disappears while the protein has it, and comes out orange - the same handoff the loading screen makes. The tab and home-screen icons were a completely different drawing; they are now the same transporter, redrawn for each size so it still reads at 16 pixels. Reduced-motion settings hold the mark still.
Each topic now shows how many questions you have left in it rather than how many it holds, and a topic you have finished greys out and says Complete. Starting a quiz serves only what you have not answered yet — pick a topic you are 6 of 12 through and you get those 6. Select All skips finished topics, so it queues up what is left instead of everything. Reset Progress in the QBank is still there if you want to take a topic again.
The Secondary Active Transport QBanks used to remember your progress in one browser only — open one on your phone and you started from nothing. Sign in and it now follows your account across devices, so the exam skins' scored palette and running percentage reflect everything you've done rather than just this laptop. Your home page also shows what you've answered in each tool. Still free, still no login needed to use them; signing in is what makes them remember you.
Some generated questions asked for a group — "which three enzymes are disrupted", "which amino acids accumulate" — but offered one member per option, so several answers were equally correct. Those are now rejected before they reach you, along with any question whose own explanation admits a wrong answer is also right. Thanks to the student who reported two of them in one sitting and worked out exactly what was going wrong.
The question palette down the left now scores itself as you go — filled green for right, hollow red for wrong, with the question you're on keeping its ring. The header adds a running "N answered · P% correct". Only the current sitting is scored, so starting a quiz doesn't open on a wall of old red.
Examplify's dark setting is now a skin of its own in every Secondary Active Transport QBank, alongside the Examplify and NBME skins — same layout, charcoal field, easier on the eyes at 1am. Pick it on the QBank's home page. Thanks to the student who suggested it.
Two kinds of unanswerable card are gone. Cards that hid the whole thing — "GH excess increases the risk of [...]" — now leave the category showing, so it reads "the risk of [...] cancer" and you know what's being asked: [...] syndrome, [...] deficiency, [...] carcinoma, and so on. And cards whose sentence gave you nothing to aim at, like "GH excess can lead to [...]", which has a dozen right answers, are no longer written at all; the deck fills the space with a real card instead.
Resuming an unfinished quiz re-opened it blank from the first question, losing every answer you'd already given. Your answers are now saved as you make them, and resuming puts them back and drops you at the first question you haven't done yet. Because they save as you go rather than when you close the tab, a closed tab, a flat battery or a crash won't cost you the quiz either.
Endocrine joins Secondary Active Transport: 372 practice questions and 124 drag-and-drop games, written from all 31 lectures of the Year 2 Endocrine block and organised the way the course runs — pituitary, thyroid, parathyroid and bone, adrenal, and the endocrine pancreas through to the diabetes emergencies. Free with any account, on your home page under Endocrine.
The eye beside each answer now works: click it to rule that option out — it's struck through and dimmed — and click again to bring it back. Ruling an option out doesn't answer the question, so you can narrow the field first the way you would in the real software. Your cross-outs clear once you answer.
A second pass on the Examplify Skin, worked from a full-resolution screenshot of the real item view: the flag pill is outlined until you flag a question and fills orange once you do, the options are full-round pills with the answer range printed above them, and the question palette fills in as you go — solid for what you've seen, a ring for where you are. The NBME Skin now keeps its item number and Mark control in their own band under the header, the way the board interface does.
Every Secondary Active Transport QBank now has an Examplify Skin and an NBME Skin — practise in an interface built to match the software you actually sit exams in, down to the layout: Examplify's left-hand question palette, flag pill and pale blue navigation bar, or the NBME item bar with its Mark checkbox. Pick a skin on the QBank's home page; it's free with any Active Transport account, so open the QBank from the hub while signed in.
If your notes contained a heading like a summary chart or a comparison table, the generator could write questions about the chart itself — what it's for, who it's aimed at — instead of the medicine underneath it. Those headings are now stripped before generation, and any question that asks about a document's formatting is rejected. Thanks to the student who reported five of them in one sitting.
While your questions generate, the membrane now spans the screen with the transporter working in it, the substrate streams rather than blinking on a rail, and the whole scene pulses as each ATP is released. It also tells you what it's doing — the live stage while a quiz builds, and a note that it's still working when a long document takes a while.
A recent change had decks covering the same concept twice in different words. Repeats are now caught three ways: when the deck's concepts are chosen, after the cards are written, and by a final pass that reads the finished deck and spots repeats even when the wording shares nothing. Removed repeats are backfilled with real cards, so your deck doesn't shrink. Thanks to the student who reported this.
Cards like "Adrenaline release [...], raises blood pressure" hid both the direction and the thing affected, so you only learned the question after flipping. Blanks now show their options wherever the answer is one of a known set — [increases or decreases], [gram-positive or gram-negative], [obstructive or restrictive], [autosomal dominant, autosomal recessive or X-linked] and many more — with a short category hint elsewhere. Hints are left off only for the headline fact a card exists to test, so a card naming a diagnosis is never reduced to picking acute or chronic.
Answer choices are shuffled for every student, but the per-choice bars were counted by position — so picks from different shufflings were pooled together and the percentages were meaningless. Picks are now matched by the answer itself. The old counts couldn't be untangled, so they've been cleared and the bars will rebuild as questions get answered; your scores and accuracy were never affected.
If the server was redeployed or restarted while your questions were still being written, the quiz stopped early — you might ask for 16 and get 7, with a "stopped loading" notice. The page now reconnects and finishes the missing questions automatically instead of giving up.
On the Anki deck-priority sliders, the recommendation button now looks like the action it is until you use it, then settles into "Recommendation applied" so you can see at a glance that the sliders already hold it. Move any slider — or change your notes — and it comes back.
Across all the study-tool QBanks (including every Picture QBank), correct answers that were noticeably longer than the other choices have been tightened — over 1,300 of them — so answer length no longer gives anything away. Wording and meaning are unchanged.
In the study-tool drag-n-drop games, X-vs-Y comparisons (Crohn vs UC, the groin hernias, acute vs chronic inflammation, and ten more) now have the things being compared as the big drop boxes, with their features sorted underneath — the way you actually think about them. Thanks to the students who suggested it.
The recommendation button on Anki generation now actually reads your notes with AI — pasted text or an uploaded PDF/TXT — and sets the priority sliders to match what your document contains.
New priority sliders on deck generation: weight diseases, tables, numbers, mechanisms and treatment from Skip to Focus, and the deck follows your mix.
Long documents were silently losing their final sections — the last learning objectives and end-of-document tables never became cards. Decks now draw cards evenly from the first page to the last.
A progress bar shows while your deck builds, and 30+ card decks no longer time out.
Questions pulled from the qbank now match on the meaning and specifics of your notes, not coarse topic labels.
Every question in the QBank now carries per-choice rationales (up from 89%), and an audit of the remainder found and repaired 261 defective questions — re-keyed, rewritten, or retired. Thank you to everyone who reported questions; most of the fixes started from your reports.
An Anki-style picker for your review queue: filter by organ system and discipline, and suspend topics you don't want yet.
Streaks now come from finishing quizzes instead.
Finish a quiz, share the link, and classmates take the same questions.
Your quiz opens at question 1 while the rest generate in the background — no more waiting for the whole set.
Question and Anki-card generation moved to the newest model generation for noticeably better clinical accuracy.
Highlighter, answer-choice strikeout, and a draggable lab-values reference panel — in quizzes and Daily Study, like the real exam interface.
Pick your cloze and explanation colors and base font size; applies to both .apkg and CSV exports.
Advanced question sets for the class study tools — Immunology, Pharmacology, BOOM, Blood & Lymph, Pathology, and GI — rolling out through early August.
Create or join public and private study groups with weekly leaderboards. Plus an in-app feedback button that goes straight to the founder.
Forgot your password? There's a proper reset flow now.
The full "active transport" visual identity — membrane motifs and all — plus mobile fixes, faster page loads, and keyboard-accessible answering.
The class's other study tools — subject QBanks, picture games, anatomy ID — collected in one place on your home page.
Join your class cohort and compete on questions answered, streaks and accuracy.
New accounts choose when to turn on spaced review instead of being enrolled automatically. Promo codes also work at checkout now.
New signups get a 5-question quiz ready to take immediately — no setup screen.
Every new account starts with a week of full Pro access — no card required.
Export the questions you missed as a ready-to-import Anki deck of cloze cards.
The site now defaults to a light look; dark mode stays one click away.
Upload an .apkg and get a personalized qbank quiz weighted toward the topics your mature cards say you know — testing what you should remember.
One-click Google sign-in alongside email and password.