The Letterbox Deadline: What the iPhone Duo Demands of Every iPhone App

Apple announced the iPhone Duo on September 9. It's the first foldable iPhone, it opens into a 7.6-inch display and closes into a 5.4-inch phone, and it starts at $1,999, with preorders on October 16 and devices in stores on October 23. Most of the coverage went to the hinge, the crease, the titanium and the weight, which is fair, because those are the things you notice when you pick one up.

Apple also published a developer tech talk that day, and for anyone running a mobile team it's the more practical document. Apple says your app runs on the Duo even without recompiling, "but screen usage improves with each SDK," and when 9to5Mac spelled out what that means for an app built with an older SDK, the answer was a letterboxed app, a phone-sized rectangle sitting in the middle of the big inner display with black around it. The people buying this device are paying at least two thousand dollars for the bigger screen and they're going to open the apps they already use on day one. Apple gave developers about six weeks between the announcement and availability, and as of this weekend the Xcode build that lets an app fill the display still hadn't been released.

What Apple shipped

A lot of the pre-launch rumors got the specifications wrong, and some of those numbers are still showing up in coverage written after the event, so these come from Apple's own spec and newsroom pages.

The Duo is a book-style foldable with a short, wide shape. Open, it measures 164.6 by 117.8 millimeters and 5.2 millimeters thick, which Apple says makes it the thinnest iPhone it has made, and closed it is 84.1 by 117.8 millimeters and 11.3 millimeters thick. It weighs 254 grams.

The inner display is a 7.6-inch folding OLED at 1878 by 2670 pixels and 430 pixels per inch with a matte nano-texture finish, and the outer display is a 5.4-inch OLED at 1398 by 2034 pixels and 460 pixels per inch. Both run 120Hz ProMotion with Always On and both reach a claimed 3,000 nits outdoors. Apple describes the inner display as 50 percent larger than the iPhone 18 Pro Max screen, and says the outer display has 90 percent of the screen area of an iPhone 18 Pro.

The chip is the A20 Pro on a 2-nanometer process, with a 6-core CPU and a 7-core GPU, and Apple claims up to 20 percent faster CPU performance than the A19 Pro and up to 40 percent faster graphics. Apple doesn't say how much RAM it has, which matters if you're trying to figure out how your app behaves with two apps open side by side. Storage comes in 256GB, 512GB, 1TB and 2TB.

The cameras are where the shape forced a compromise. There's a 48MP main camera and a 48MP ultra wide, and Apple lists a 12MP "optical-quality 2x Telephoto" that shares the main camera's f/1.6 aperture and that Apple describes as integrated, like having two cameras in one, so there is no separate telephoto lens. There's a 12MP camera on the outer display and an under-display camera on the inner one.

Authentication changed too, since the Duo uses Touch ID in the side button and has no Face ID. Apple didn't explain that on stage. My guess is that a 5.2-millimeter half leaves very little room for a TrueDepth camera module, but that's my inference and not anything Apple has said.

Apple claims up to 24 hours of typical use from dual built-in batteries, and up to 44 hours of video playback on the outer display or 31 hours on the inner one, without publishing capacity in milliamp-hours. Charging reaches 50 percent in about 20 minutes with a 60-watt or larger adapter, MagSafe and Qi2 wireless charging go up to 25 watts, the modem is Apple's own C2 with mmWave in the United States, and the phone is eSIM only everywhere.

The crease

For about two years the most repeated rumor about this phone was that it would have no crease, and Apple doesn't claim that. The inner display is made of 10 ultrathin layers, optically clear adhesives "let the display layers glide as the screen folds, minimizing stress and crease," high-strength glass sits above and below the folding panel with a titanium plate behind it, and the matte nano-texture finish "minimizes crease visibility." Apple uses minimize both times and doesn't publish a crease depth.

The hinge has more than 100 components, an integrated magnet array and a 3D-printed hinge cover made from recycled titanium, inside a Grade 5 titanium frame. The liquid-metal hinge that got so much pre-launch attention isn't mentioned anywhere in Apple's materials.

Early impressions don't agree on how visible the crease is. Engadget's Sam Rutherford described it as very faint and easier to feel with a finger than to see, while 9to5Google's Andrew Romero, going off hands-on videos from the event, wrote that it looks just as visible as it does on Samsung's Galaxy Z Fold8. How much of a crease you see depends a lot on the angle and the lighting, so I'd wait for reviews from people who have used it for a few weeks before deciding who's right.

Apple also hasn't published a fold-cycle rating for the Duo, and as of this weekend there was no AppleCare+ price and no out-of-warranty repair pricing for the screen or the hinge. The folding display flexes every time the phone opens, so if you're thinking about buying these for a team, I'd wait for those numbers before putting in a purchase order.

Weight and thickness

At 254 grams and 11.3 millimeters closed, the Duo is heavier and thicker than the Galaxy Z Fold8, which Samsung lists at 201 grams and which measures 9.7 millimeters closed, so the difference is 53 grams.

Some of that weight is easy to account for from the spec sheet, which lists a titanium frame and hinge, two batteries, a vapor chamber for sustained performance, and glass on both sides of the folding panel. Whether all of that makes it more durable than Samsung's lighter phone is exactly what a fold-cycle rating would tell us, and Apple hasn't given one. Apple is usually late to a category and usually doesn't top the spec comparison when it arrives, and what it tends to bring instead is one company controlling the chip, the modem, the operating system and the developer tools on the same schedule.

How existing apps behave on the Duo

In its "Prepare your app for iPhone Duo" tech talk, Apple says an app runs on the Duo even without recompiling, "but screen usage improves with each SDK." An app built with the iOS 27 SDK extends to the left of the status bar on the inner display, and one built with the iOS 27.1 SDK reaches the screen edge. 9to5Mac's write-up put it in plainer terms:

• Built with an SDK older than iOS 27: the app runs letterboxed inside a safe area on a black background.
• Built with the iOS 27 SDK: the app uses more of the display, but with gaps.
• Built with the iOS 27.1 SDK: the app fills the display.

An app that hasn't been rebuilt still works and nothing crashes, it just shows up as a phone-sized box on a much bigger screen.

The same tech talk gets specific about the code, and most of it is work Apple has been nudging developers toward for years, with a foldable now attached to it:

• Use size classes, not orientation. The inner display reports a regular width and regular height size class, the same combination iPad apps already handle, and Apple says the inner display doesn't honor an app's supported interface orientations, so layout logic that assumes portrait means phone will be wrong.
• Move off UIScreen.main. Apple says main-screen references are ambiguous on this device and will be deprecated, and recommends getting the screen from the window scene or reading displayScale from the trait collection.
• Handle asymmetric safe areas. Insets can differ on each side of the device.
• Prefer the system navigation containers. NavigationSplitView and UISplitViewController adapt between compact and regular layouts, and tab bars can move to a sidebar placement on the larger display.
• Keep content away from the fold. The iOS 27.1 SDK adds ReservedRegion in SwiftUI and UIViewReservedRegion in UIKit, so content can avoid the center of the display when the device is partly folded.

Apple also says Xcode's DeviceHub lets developers open, close, rotate and fold a simulated device, so teams can test the different postures without the hardware.

What WWDC already required

Teams that went through the WWDC sessions in June saw part of this coming. In "Modernize your UIKit app," Apple said the UIScene lifecycle is now required when building with the latest SDKs and that without it an application will no longer launch. It also changed UIRequiresFullscreen so it no longer fully opts an app out of resizing, and made iPhone apps fully resizable on iPad and in iPhone Mirroring.

That session never mentioned a foldable, but an app that adopted scenes, used size classes and handled resizing back in June is already most of the way to running well on the Duo.

What the April 2027 rule does and doesn't do

Apple has also announced that starting in April 2027, apps submitted to the App Store must be built with the iOS 27 SDK or later, and it would be easy to read that as Apple eventually forcing everyone onto the new layout. It won't do that. Building with the iOS 27 SDK moves an app from letterboxed to extending past the status bar, while filling the display takes the 27.1 SDK plus layouts that actually work at a regular width and regular height, and that second part is design and engineering work no SDK requirement does for you.

I'd expect apps from teams that already ship on iPad and already use adaptive layouts to adapt fairly quickly. Phone-only apps will take longer, and that includes a lot of internal enterprise apps and apps on cross-platform frameworks that have to wait for their framework to support the new SDK first.

The iPad precedent

When the original iPad shipped in 2010, apps built only for iPhone ran on it in a small window in the middle of the screen, with a button to scale them up to double size. They worked, and they looked like software built for a different device.

The apps iPad buyers cared about most got rebuilt within months, and whether an app had a real iPad version became something reviews and rankings paid attention to, while a long tail of apps stayed in the small window for years and some never left. Apple didn't have to force it, because customers had paid for a bigger screen and noticed which apps used it.

The Duo sets up something similar, with two differences. The tooling is much better now, since size classes, scenes and split view containers didn't exist in 2010 and a well-structured SwiftUI app can adapt without much code. And the buyers are more concentrated, because the original iPad was a new category at $499 while the Duo is an upgrade at $1,999 and up, which means almost everyone who buys one is an existing iPhone customer who already has your app installed. That makes the audience small in absolute numbers, which is a legitimate reason to weigh it against other work, and it also means those customers will be opening your existing app on day one.

That chart is an estimate, not survey data. I think the best early signal for how much work the Duo will be is whether the team already treats iPad as a real target.

A plan for the next six weeks

For most consumer apps the practical question is how to fit this work in over the next six weeks alongside everything already on the roadmap. Here's the order I'd take it in.

1. Audit for the known failure modes this week. Search the codebase for UIScreen.main, for orientation locks the inner display will ignore, for hard-coded widths and phone-only layout assumptions, and for apps still on the pre-scene app delegate lifecycle. Each of those is a concrete ticket and none of them needs hardware.

2. Decide which tier each app should hit on October 23. For a flagship consumer app, letterboxed probably isn't acceptable, the iOS 27 tier with gaps may be a reasonable place to be at launch, and filling the display with a proper regular-width layout is where you want to end up. It helps to write down which app lands where, so the tradeoffs are visible.

3. Reuse your iPad work. If you ship an iPad version, the regular-width layouts, split views and sidebars already exist, and since the Duo's inner display is close to an iPad mini in size, a lot of that code should carry over.

4. Design for sizes instead of a special Duo mode. Apple's design talk recommends designing for compact and regular sizes rather than a custom layout for each posture, keeping controls along the edges, and letting system components move interactive elements away from the center when the device is partly folded, which means the work usually goes into a better adaptive layout rather than a Duo-specific screen.

5. Plan for Split View at half width. iOS 27 brings Split View to iPhone for the first time on the Duo, so two apps can run side by side. MacRumors reports it's a fixed 50/50 split, which would make each half roughly the width of the outer display, although Apple hasn't published the details and Mac Observer points out Apple hasn't said whether the divider can be resized. Either way your compact layout is going to run next to another app more often than it does today.

6. Get the 27.1 toolchain into CI as soon as it ships. The Xcode 27.1 beta was listed as coming later this month and hadn't been released as of September 13. The time between that beta and a production release is the window you actually get, so it's worth lining up the build pipeline, DeviceHub simulation and screenshot tests now.

What the code change looks like

For SwiftUI teams, most of the change is about letting the size class decide the layout instead of the device model, and using containers that already collapse and expand.

struct LibraryView: View {
    @Environment(\.horizontalSizeClass) private var horizontalSizeClass

    var body: some View {
        NavigationSplitView {
            CollectionList()
        } detail: {
            if horizontalSizeClass == .regular {
                // Open Duo, iPad, or a wide window: show richer detail
                ExpandedDetailView()
            } else {
                // Closed Duo, a standard iPhone, or half-width Split View
                CompactDetailView()
            }
        }
    }
}

NavigationSplitView already collapses into a single stack at compact width and expands into columns at regular width, so the same view tree works for the closed phone, the open display and an app running at half width next to another app.

In UIKit codebases, the change that tends to show up most often is removing global screen assumptions. Asking the process-wide main screen for its bounds or scale made sense when an iPhone had one display and one window, and it stops making sense on a device with two displays and multiple windows.

// Before: a global assumption that breaks on a two-display device
let scale = UIScreen.main.scale

// After: ask the environment the view actually lives in
let scale = traitCollection.displayScale
let screen = view.window?.windowScene?.screen

Each of these is small on its own. The effort comes from how many places a mature codebase made the same assumption and from the testing it takes to check every screen at compact and regular widths, which is why the audit in step one helps, because it turns "support the Duo" into a list of call sites you can count and estimate.

Cross-platform frameworks

Teams on React Native, Flutter or similar frameworks have an extra dependency. The SDK tiers depend on the iOS SDK the final app binary is built with, so the framework's iOS build tooling has to support the iOS 27.1 SDK before an app can fill the display. After that the layout work is the same idea, responding to window size instead of assuming a phone, using APIs like useWindowDimensions in React Native or MediaQuery in Flutter.

This week I'd check your framework's release notes and issue tracker for iOS 27.1 and Duo support and plan on some lag between Apple's toolchain and your framework's. If that lag looks long, at least make sure the app builds cleanly against the iOS 27 SDK, which gets it out of the fully letterboxed tier.

Migrations like adopting scenes or supporting resizing tend to look optional right up until hardware ships that depends on them, and it's usually cheaper to take them a piece at a time than all at once against a launch date.

Price and memory costs

Reports before the event had pointed to a price of $2,300 to $2,500, and Apple came in at $1,999 for the base model, about $100 above the Galaxy Z Fold8 and the Pixel 11 Pro Fold.

Wall Street's reaction was mixed. Apple shares slipped a little on the day of the event and then rose about 3.6 percent the next day to close near $326.57, Bank of America kept its Buy rating while trimming its price target from $380 to $370, and Rosenblatt held a Neutral rating with a $303 target. I think Apple priced it to sell more of them in the first year, even if that means making less on each one.

That fits what's happening with memory prices. IDC forecast in late August that global smartphone shipments would fall 16.7 percent in 2026, a record drop it attributed to the memory crisis. I made a related call in December, when I predicted that AI-driven memory shortages would push consumer device prices up 15 to 25 percent by the third quarter of 2026, and that prediction comes up for grading at the end of this month with the fall launches as part of the evidence. The iPhone 18 Pro starting price went up $100 to $1,199. Apple didn't attribute that to memory and one price point doesn't prove anything, but it's the direction I predicted.

When the overall market is shrinking, an expensive phone aimed at people who care least about price increases is a reasonable bet, and it gives Apple's biggest spenders a new reason to upgrade in a year when a lot of people are holding onto the phones they have.

For enterprise buyers

If you buy phones for a team, a few details on the spec sheet matter more than they did in the keynote.

Authentication. The Duo uses Touch ID in the side button instead of Face ID. Most enterprise mobile policies and single sign-on flows go through the same system biometric prompt either way, so apps using the standard authentication APIs shouldn't notice, but custom flows, kiosk-style deployments and help-desk documentation written around Face ID will need a look.

eSIM only, in every market. There's no physical SIM slot anywhere, so an organization still provisioning corporate lines with physical SIM cards will need an eSIM process with each carrier. A lot of fleets have already moved to eSIM, but it's worth checking before a pilot.

Repair and coverage. Without an AppleCare+ price or out-of-warranty screen and hinge repair pricing, you can't model the total cost of ownership yet, and for a display that folds many times a day, repair is likely to be the biggest variable in a three-year cost.

Who it's for. A 7.6-inch display with two apps side by side is useful for people who work across email, documents, dashboards and messaging all day, and a lot less useful for roles that mostly take calls or scan barcodes. I'd start with a small pilot of the people most likely to benefit, compared against their current phone and tablet, once repair pricing is public.

The foldable market

Foldables are still a small category. Omdia counted 18.2 million foldable shipments in 2025, up 6 percent, and Counterpoint puts foldables at about 2 percent of all smartphones. Samsung has sold book-style foldables since 2019 and Huawei has gone further with tri-folds, and neither has made them a mass-market product. The 2026 forecasts, published just before and after the Duo launch, lean heavily on Apple.

TrendForce expects the Duo to take nearly a quarter of the foldable market in its first year at about 5 million units, tied with Huawei and behind Samsung. Counterpoint puts Apple at 25 percent, and IDC expects Apple to reach roughly 40 percent of foldables in 2027, the first full year the phone is on sale.

Estimates for the Duo itself are all over the place, mostly because nobody knows how fast production will ramp.

These three numbers measure different things. TrendForce and Counterpoint estimated calendar 2026 shipments in September, which covers a little over two months of sales, and Nikkei reported in July, before the name or price was public, that Apple had asked suppliers to prepare about 10 million units this year. The distance between the build plan and the shipment estimates suggests production, rather than demand, will decide the 2026 number.

Even 5 million units is a small slice of the active iPhone base. For app teams, what matters more is who those 5 million people are, which is people who spent at least $1,999 on a phone and who are more likely than most to pay for subscriptions, use productivity apps for work and leave reviews.

Open questions

Supply-chain reporting has already started filling in the things Apple didn't say, with numbers that sound more certain than they are.

• Display supplier. Korean press reports that Samsung Display is the sole supplier of the foldable panel, which Apple hasn't confirmed, and a widely repeated per-panel price traces back to a single leaker.
• Bill of materials. There's no teardown-based cost estimate yet, so any precise component cost you see right now is a guess.
• RAM and battery capacity. Apple doesn't disclose either, and the figures going around come from pre-launch reports.
• Durability and repair. No fold-cycle rating, no AppleCare+ price for the Duo and no out-of-warranty repair pricing as of this weekend, and iFixit has already said it's skeptical the phone will be easy to repair.
• Apple Pencil. Apple says the Duo will support Apple Pencil with USB-C, and lists that as coming later this year.

I'd wait for teardowns, repair pricing and a few months of real use before buying fleets of them or believing anyone's durability claims.

Siri and Apple Intelligence

The new Siri starts rolling out as a beta with iOS 27 on September 14 for supported devices set to English, with more languages in October, and it isn't initially available in the European Union. Duo owners get it when their phones arrive.

That Siri runs on Apple's arrangement with Google, which I covered in the WWDC 2026 analysis of Apple's billion-dollar Gemini dependency, and in China Apple went a different way, which I wrote about in the piece on Apple shipping Alibaba's Qwen models as a regional model layer. A bigger display gives an assistant more room to show what it's doing, and two apps on screen give it more to work with, although at launch the work app teams face on the Duo is mostly layout work.

What to watch

Over the next six months a few things will show whether the Duo turns foldables into a mainstream category or stays an expensive niche.

On the market side, Apple's fiscal first-quarter iPhone results in early 2027 will be the first real read on demand, and by then we'll have a better sense of whether IDC's forecast of roughly 40 percent of foldables in 2027 looks high or low. For more on how the market has been pricing Apple this year, see my news analysis of the Nvidia and Apple market-cap crossover in July.

Where this leaves app teams

The Duo is late to the category, heavier and thicker than Samsung's closest phone, and priced lower than most people expected, and Apple has left real gaps, with no crease-free claim, no fold-cycle rating and no repair prices yet.

For people who build apps, the date that matters is October 23, when owners start opening the apps they already use on the inner display. Apple has been clear about what decides how those apps look, and the work is familiar: scenes, size classes, adaptive navigation, no global screen assumptions, and layouts that keep content away from the fold. If your team hasn't started on it, the audit in step one is a reasonable place to begin this week.