AutoSleep

Sleep tracking has become the subscription industry’s favorite racket. Eight Sleep wants thousands for a mattress cover. Oura wants a monthly fee on top of a ring. Even software-only players like Sleep Cycle and Rise have migrated to recurring billing. Against this backdrop, AutoSleep — built by tiny Sydney developer Tantsissa — feels almost anachronistic: a one-time $8.99 purchase that has been continuously updated since before Apple even shipped native sleep tracking.

The pitch is disarmingly simple. Wear your Apple Watch to bed. That’s it. No start button, no bedtime mode to remember, no ritual. By morning, a notification hands you a full breakdown: time asleep, sleep quality, deep sleep, cycles, heart rate dips, and blood oxygen. The app’s heuristics do the detection work that Apple’s own Sleep app still demands scheduling for.

And the market has noticed. AutoSleep holds a 4.7/5 rating across more than 61,000 App Store ratings. More telling: it offers actigraphy exports rigorous enough for sleep clinics and universities to use in validation studies. When academics trust a consumer app’s raw data, that’s something more serious than a wellness toy.

The catch — and there is one — is that AutoSleep’s years of feature accretion have produced an interface that can feel like mission control when you just wanted to know if you slept well. More on that below.

• Fully Automatic Detection: No buttons, no schedules, no “I’m going to bed now” taps. The app’s heuristics infer sleep from Watch sensor data, which means the nights you forget about are still captured. This is the feature everything else hangs on.
• Sleep Apnea & SpO2 Tracking: Recent updates added apnea tracking with gauge and trend graphs, plus support for third-party devices that push blood-oxygen data into Apple Health — a genuinely useful bridge for people juggling multiple sensors.
• Actigraphy Export: Research-grade movement data exports, rare among Watch sleep apps. If you’re working with a sleep clinic, this turns a consumer gadget into a legitimate data source.
• On-Device Privacy: No analytics, no ads, no server uploads, no third-party code. Your sleep data — arguably among your most intimate health information — never leaves your phone.
• Smart Alarm: Wakes you at the optimal point in your sleep cycle rather than at an arbitrary minute, within a window you define.

The Loop That Isn’t a Loop

The strength of AutoSleep is that its core interaction is nothing. Most health apps fail because they demand a habit; AutoSleep only demands that you keep doing something you already do — wearing your watch. This eliminates the single biggest failure mode of sleep tracking: the missed night. Competitors that require arming a bedtime mode inevitably develop gaps in their data, and gappy sleep data is nearly worthless for spotting trends. AutoSleep’s dataset is continuous by default, and that continuity is what makes its trend explorer, weekly averages, and sleep bank actually meaningful rather than decorative.

The sleep bank deserves particular mention. Rather than judging each night in isolation, it treats sleep like a ledger — deposits and withdrawals against your goal — which maps far better to how sleep debt actually works physiologically. A bad Tuesday matters less if the week balances out. It’s a smarter mental model than the nightly pass/fail grade most apps issue, and it subtly discourages the anxiety-inducing perfectionism that sleep trackers can breed.

The Clock Face: Clever, Then Crowded

Sessions render on a clock face that shifts from red to green as you approach your sleep goal. As an at-a-glance readout, it works — one look tells you whether last night was a win. The problem arrives when you go one level deeper. Rings, wedges, dots, and overlapping arcs accumulate until the display starts resembling an aviation instrument. Tantsissa clearly knows what every element means. New users won’t, and the onboarding — while improved over the years — still can’t fully paper over years of accumulated UI concepts.

This is the app’s honest weakness: the interface is dense and initially confusing. It’s the classic indie-app trade-off. A single passionate developer keeps adding capabilities users request, and each one is individually defensible, but nobody ever gets hired to say no. AutoSleep needs an editor more than it needs another feature.

Accuracy and the Heuristics Question

AutoSleep predates Apple’s native Sleep Stages and now integrates with them rather than fighting them — the right call. Its own heuristic analysis of deep sleep and cycles runs alongside Apple’s staging, which gives you two perspectives on the same night. Purists will note that wrist actigraphy plus heart rate is not polysomnography, and no Watch app can truly measure sleep stages the way a lab can. AutoSleep, to its credit, doesn’t oversell this; the actigraphy export exists precisely so researchers can validate against clinical instruments. That intellectual honesty is rare in a category thick with pseudo-scientific “sleep scores.”

The Privacy Position

The 100% on-device architecture isn’t marketing garnish — it’s a structural decision with real consequences. There’s no account to create, no cloud sync to breach, no data broker downstream. Period trackers and fitness apps have faced subpoenas and data-sale scandals, so keeping sleep data (which can reveal insomnia, apnea, shift work, and more) entirely local is a feature worth paying for on its own. The trade-off is equally real: lose your phone without a backup and your history goes with it, and there’s no web dashboard for reviewing data on a bigger screen.

Aesthetically, AutoSleep is functional rather than fashionable. The color-coded clock face is its best idea; the deeper screens lean on dense grids of numbers, gauges, and small-multiple charts that prioritize information over elegance. It looks like software built by an engineer who loves data — because it is. Compared to the airy, whitespace-heavy design of Apple’s own Sleep app, AutoSleep feels like the difference between a dashboard and a brochure. Some users will prefer exactly that.

Performance is, by design, a quiet strength. Because analysis happens on-device with no network round-trips, there’s no server dependency standing between you and the morning breakdown. The app’s requirements are refreshingly modest — watchOS 6+ and iOS 14+ — meaning owners of aging hardware aren’t forced into an upgrade cycle just to keep tracking.

AutoSleep is iOS-only in the truest sense: it requires both an iPhone and an Apple Watch, full stop. No Android version exists, and none is plausible given the app’s total dependence on Watch sensors. The Watch itself does the sensing while the iPhone app does the heavy analytical lifting — a sensible division of labor. Widgets bring key metrics to the iPhone home screen, and the Apple Health integration is deep and bidirectional: AutoSleep reads SpO2 from third-party devices via Health and writes its sessions back, so it plays well in a multi-app health stack rather than hoarding its data. If you sleep with your Watch charging on the nightstand, this app is not for you — a charging-routine adjustment is the real onboarding cost.

AutoSleep is what happens when a developer spends years sanding one product against the same problem: it’s deep, trustworthy, and priced like software used to be. Its flaws are the flaws of abundance — too much data, arranged too densely — not of laziness or cynicism. Apple’s built-in Sleep app is simpler and free, and for casual users it’s enough. But the moment you want to understand why you’re tired, spot a trend, hand real data to a doctor, or simply refuse to rent your own sleep history back from a subscription service, AutoSleep is a strong pick on Apple Watch. It’s a solid, serious utility with real rough edges — recommended for the right audience, not for everyone.