Hello,
After upgrading to macOS 26.2, I’ve noticed a significant performance regression when calling evaluateJavaScript in an iOS App running on Mac (WKWebView, Swift project).
Observed behavior
On macOS 26.2, the callback of evaluateJavaScript takes around 3 seconds to return.
This happens not only for:
evaluateJavaScript("navigator.userAgent")
but also for simple or even empty scripts, for example:
evaluateJavaScript("")
On previous macOS versions, the same calls typically returned in ~200 ms.
Additional testing
I created a new, empty Objective-C project with a WKWebView and tested the same evaluateJavaScript calls.
In the Objective-C project, the callback still returns in ~200 ms, even on macOS 26.2.
Question
Is this a known issue or regression related to:
iOS Apps on Mac,
Swift + WKWebView, or
behavioral changes in evaluateJavaScript on macOS 26.2?
Any information about known issues, internal changes, or recommended workarounds would be greatly appreciated.
Thank you.
Test Code Swift
class ViewController: UIViewController {
private var tmpWebView: WKWebView?
override func viewDidLoad() {
super.viewDidLoad()
// Do any additional setup after loading the view.
setupUserAgent()
}
func setupUserAgent() {
let t1 = CACurrentMediaTime()
tmpWebView = WKWebView(frame: .zero)
tmpWebView?.isInspectable = true
tmpWebView?.evaluateJavaScript("navigator.userAgent") { [weak self] result, error in
let t2 = CACurrentMediaTime()
print("[getUserAgent] \(t2 - t1)s")
self?.tmpWebView = nil
}
}
}
Test Code Objective-C
- (void)scene:(UIScene *)scene willConnectToSession:(UISceneSession *)session options:(UISceneConnectionOptions *)connectionOptions {
NSTimeInterval startTime = [[NSDate date] timeIntervalSince1970];
WKWebView *webView = [[WKWebView alloc] init];
dispatch_async(dispatch_get_main_queue(), ^{
[webView evaluateJavaScript:@"navigator.userAgent" completionHandler:^(id result, NSError *error) {
NSTimeInterval endTime = [[NSDate date] timeIntervalSince1970];
NSLog(@"[getUserAgent]: %.2f s", (endTime - startTime));
}];
});
}
