ANTIGRAVITY LABJP
Articles/App Development
App Development/2026-07-18Intermediate

Which Device Gets the iOS 27 Public Beta? Choosing a Test Phone When There's No Way Back

The iOS 27 public beta arrived on July 13 carrying the same build number as developer beta 3. Starting from that one detail, here is how I decide which device gets it, what to archive first, and where to look in my own app.

iOS 273public betatest deviceindie development16Liquid Glass

On the morning of July 13, the iOS 27 public beta announcement showed up, and the first thing I checked was the build number. 24A5380h. The same string already sitting on my test device as developer beta 3.

That settled it. There was no work to do.

It was just a build number matching, but beta season is mostly made of small checks like this one — confirm, then move. Maintaining apps on my own means the devices and hours available for testing are finite. Which is exactly why a few things have to be decided before anything gets installed.

What a matching build number tells you first

The public beta shipping as 24A5380h is the identical build to developer beta 3. Different label, same bits.

The read is simple. If a device is already on the developer beta, moving it to the public beta gains you nothing, because there is nothing new on the other side.

If you have not installed anything yet, though, the public beta is now the shortest path to iOS 27. You reach the same bits without routing through developer account setup.

Where you areWhat to do as of July 13
Developer beta 3 is installedNothing. Same build, so switching is a no-op
No beta yet, and you own a spare devicePut the public beta on the spare
No beta yet, and you own one deviceSkip it — reasoning in the next section

What's inside: new Apple Intelligence capabilities, a stronger Siri, new parental controls, Liquid Glass refinements (there's now a slider for how strong the effect looks), Spotlight improvements, and general performance work.

Of that list, the one that lands directly on your app's appearance is the Liquid Glass intensity slider. I'll come back to why.

Pick the device before you install anything

The hardest thing to undo in beta season isn't the OS. It's the device you chose to put it on.

Signing for older iOS versions eventually closes, and once it does there is no supported route back. On top of that, a backup taken on the newer OS won't restore onto the older one. So "install it, and roll back if it goes badly" tends to be a one-way trip in practice.

Three lines I hold:

  • Nothing goes on the device carrying my daily messages and payments
  • Nothing goes on the device I use for final on-device purchase-flow checks
  • The beta device never signs in with my primary working accounts
Device roleiOS 27 beta?Why
Daily driverNoWith no way back, the personal-life risk is too large
Production-equivalent purchase testingNoIt muddies triage for the app that's actually shipping
Dedicated older test deviceYesDamage stays inside something you can wipe and rebuild

With a single device, I come down on the side of not installing. The pull to try it is real, but when a support message arrives about the live app, a beta phone leaves you unable to separate "app bug" from "beta OS behavior." An environment you can't separate isn't a test environment.

What to archive before it goes on

Once I've decided a device gets the beta, three things get saved first.

A backup taken on the current OS. Anything captured after the beta lands won't restore to where you were. I keep an encrypted backup of the moment before.

The artifacts for the build currently on the App Store. The .ipa and the dSYM for exactly what's live. If you find a crash under the beta, a log you can't symbolicate is barely a lead.

A full set of screenshots on the current OS. This is the one that pays off most this cycle. An intensity slider for Liquid Glass means the same screen can look different on the same device. Without a reference shot, you can't tell whether the beta changed something or whether it always looked that way.

Separately: the requirement that uploads to App Store Connect be built with the iOS 26 SDK or later has been in force since April 28. That's a different thread from beta testing, and keeping them apart saves confusion. Whether your submission pipeline already satisfies it is a check to finish before beta work, not during it.

Where to look in your own app

I don't have the hours to sweep every feature on a beta device. I guess at the likely damage and order the work.

The app I maintain deals in background imagery, and it leans on one structural pattern: a translucent panel over a photo, with text on top of the panel. The Liquid Glass intensity slider lands right there. Turn the intensity down and the photo behind the panel carries more detail; turn it up and it carries less — and the contrast of the text above it moves with it.

So my check starts at the slider's minimum and its maximum, asking only whether the text is still readable. Both ends, not the middle, because if it breaks, it breaks at the ends.

iOS 27 changeWhere to look in your appHow I rank it
Liquid Glass intensity sliderAny screen with text over a translucent panelHigh — a user setting now changes your rendering
Spotlight improvementsApp Intents and anything you indexMedium — it's the side where exposure can grow
New parental controlsAge-related flows and surfacesDepends on your audience
Apple Intelligence and SiriIntents you exposeSkip it if you expose none

How far to lean on the OS-side AI inside your own app is a design question rather than a beta-testing one. I worked through that boundary in Delegate to the OS AI or Own It: Drawing the iOS 27 Feature Boundary.

Fold the routine into the same shape every time

Betas keep coming. Improvising the check each round means the set of things you look at drifts, and drifted checks can't be compared later.

I keep the checklist inside the repository — device roles, what to archive, the order of screens to review, in plain text. Each beta, I open it, and when I'm done I add one line about whatever surprised me.

Parts of it automate cleanly. But I keep the delegated scope narrow.

TaskDelegate to an agent?Why the line sits there
Walk every screen and capture screenshotsYesFixed procedure, and the result lands as files
Diff those images against the current-OS setYesListing what moved is machine work
Judge whether a diff is acceptableNoReadability needs a human looking at it
Decide which device receives the betaNoNo reason to delegate an irreversible choice

Anything whose inputs live entirely inside the repository, I hand off. Anything that lives outside it — devices, appearance — stays with me. That's the whole rule.

On distribution, App Store Connect now accepts TestFlight builds made with the Xcode 27 beta, which is the lowest-friction path for getting builds onto a test device. I wrote up automating that pipeline in Automate TestFlight Beta Distribution with Antigravity: A Practical Guide.

Official design kits for Figma and Sketch are also available now for iOS 27, iPadOS 27, and macOS 27. Sizing up the damage in design before touching hardware is a legitimate option.

The next step

If developer beta 3 is already installed, there's nothing to do this round. Same build, so standing still is the correct move.

If you haven't installed yet, the first decision is the device — not the OS. Before it goes on, ask yourself whether you could wipe that phone and rebuild it from scratch. If the answer is no, it isn't a test device.

I once assumed I could roll back, and paid for the assumption. If this saves someone that detour, I'm glad.

Share

Thank You for Reading

Antigravity Lab is ad-free, supported entirely by members like you. We publish practical guides daily with implementation code, benchmarks, and production-ready patterns. If you've found it useful, we'd love to have you on board.

  • Copy-paste ready implementation code
  • New advanced guides published daily
  • $5/mo or $10 for lifetime access
View Membership →

If you found this article helpful, a small tip ($1.50) would mean a lot to us. Your support helps keep this site ad-free and covers server and hosting costs.

Related Articles

App Dev2026-06-14
How Far Can On-Device Inference Stay Free? Measuring the Line Between Foundation Models and Gemini
WWDC 2026 widened Apple Foundation Models' free tier, making on-device inference easier to reach for. But 'free, so everything on-device' hits cases where quality falls short. Here is how to decide the range you hold on-device and the range you pass to cloud Gemini by measurement, not guesswork.
App Dev2026-07-01
Systematize Pre-Release Checks for Sound-Playing Apps with Waveform Rendering and Agents
The 6/26 update let Antigravity render audio files. Drawing on years of shipping sound in healing apps, here is a design that folds loudness, clipping, length, and silence checks into an agent verification loop, with the actual scripts.
App Dev2026-06-21
A Few Low-Density Phones Lost Their Bundled Wallpaper — The drawable vs nodpi Boundary in Play's Density Splits
App Bundle density splits will happily split images that should never be split, dropping a static resource on one density bucket only. Here is how I reproduced it with bundletool and fixed it by moving to drawable-nodpi or disabling density split — with the decision criteria.
📚RECOMMENDED BOOKS
Build a Large Language Model (From Scratch)
Sebastian Raschka
LLM Dev
Prompt Engineering for LLMs
Berryman & Ziegler
Prompting
AI Engineering
Chip Huyen
AI Eng
* Contains affiliate links
See all →