Did you know you can customize Google to filter out garbage? Take these steps for better search results, including adding my work at Lifehacker as a preferred source.The iPhone 17 series has left the realm of mere rumor. Apple announced its four new iPhones at yesterday's "Awe Dropping" event—the iPhone 17, iPhone Air, iPhone 17 Pro, and iPhone 17 Pro Max. All four use the company's new A19 chip, though different models come equipped with different versions of it: The iPhone 17 uses the base model A19, while the iPhone 17 Pro and 17 Pro Max use the A19 Pro. The iPhone Air has an A19 Pro, but it's a "binned" chip. There's a lengthy explanation of what that means, but in short, the Air uses a weaker version of the A19 Pro, so it isn't quite as powerful as the version you'll find on the Pros.While Apple frequently touts how powerful its newest iPhones are, they're never all that specific about it, especially during their big keynote announcements. The company likes to compare the new iPhones to random older iPhones, like saying the iPhone 17's A19 chip is 1.5 times faster than the A15 Bionic in the iPhone 13. But Apple doesn't usually put up the raw numbers to compare performance directly. For that, we turn to benchmarking.Benchmarking tests offer a standardized way of testing a chips' CPU (central processing unit) and GPU (graphics processing unit), and comparing their performance against other chips. Geekbench is a popular option for benchmarking, and tests a few different things: the performance of tasks that require one core of the CPU; performance of tasks that require multi-cores; and the "Metal" test, that benchmarks GPU performance.When you run a benchmark through Geekbench, your results are posted to the site's database. That means whenever a new device is announced, it's only a matter of time before benchmarking data appears there for those devices. As GSMArena reports, it seems some A19 and A19 Pro tests have already been run. These tests give us an early look at the performance of the new chips, and an idea of how much Apple has improved them over previous generations—at least on paper. As the phones, and iOS 26 itself, are not officially out yet, take these numbers with a grain of salt. With that said, here are the preliminary results.iPhone 17The A19, again found in the iPhone 17, scored a 3,608 in the single-core test, and 8,810 in the multi-core test. That represents an 8% jump over the iPhone 16's single-core score of 3,317, and a 7% jump over its multi-core score of 8,198. The A19's GPU scored a 37,014 on the Metal test, a more than 33% increase when compared to the iPhone 16's score of 27,702. The test results show that the iPhone 17 has 8GB of RAM.iPhone AirThe iPhone Air runs an A19 with a six-core CPU and a five-core GPU—one less than the iPhone 17 Pro, as this is a binned chip. GSMArena reports that it scored a 3,674 in the single-core test and 8,824 in multi-core. That's not much higher than the standard A19, though the results show Apple included four extra gigabytes of RAM in the Air over the iPhone 17.Turning to the GPU, the Air scored a 37,743 on the Metal test. That represents an improvement of just under 2% over the A19. While you'll likely notice some gains if you're coming from the iPhone 16, or even the 16 Pro (more on that below) the Air isn't going to outperform the 17 much—even with its binned A19 Pro. iPhone 17 ProWhile the iPhone Air's binned A19 Pro chip seems barely improved over the iPhone 17's A19 chip, the 17 Pro starts to show some advantage—at least in GPU. The A19 Pro in the iPhone 17 Pro scored 3,523 in single-score and 9,028 in multi-core. That's a lower single-core score, and only a slightly higher multi-core score, which is certainly interesting. In theory, the lower-performing A19 and A19 Pro chips could yield slightly faster results in single-core tasks, while the Pro could beat the bunch a bit in multi-core tasks.However, things get more interesting when it comes to the A19 Pro's GPU—The chip achieved a Metal score of 44,342, 17% over the iPhone Air, and nearly 20% better than the iPhone 17. That's not all that surprising, seeing as the Pro has six GPU cores compared to the Air's five. Compare all that to the iPhone 16 Pro's scores of 3,447 in single-core, 8,576 in multi-core, and 32,673 in Metal. It's only a 2% bump in single-core and 5% in multi-core, but more than 35% in Metal. Not a bad gain. iPhone 17 Pro MaxIf you want the best possible performance in an iPhone, the Pro Max is usually the model to get. It seems the 17 Pro Max is no exception: According to these initial Geekbench results, the 17 Pro Max's A19 Pro chip received a 3,781 single-core score, 9,679 multi-core score, and 45,657 Metal score. Those, of course, are the highest scores of any of the iPhone 17 series. The single-core score is not dramatically higher than the other phones, and the multi-core score isn't a huge jump: The most dramatic difference there is between the iPhone 17 and the 17 Pro Max at just under 9%. That said, the Metal score is pretty impressive. It's less than 3% better than the 17 Pro, but it's nearly 21% higher than the Air, and over 23% higher than the iPhone 17. We can see improvements from the iPhone 16 Pro Max's 3,430 single core, 8,502 multi-core, and 32,665 Metal scores. That's 10% improvement in single-core, nearly 14% in multi-core, and nearly 40% in Metal. There's no denying the 17 Pro Max is the phone to beat when it comes to graphics processing potential. You can look up any iPhone model you want on Geekbench's database and compare its scores to the iPhone 17 series' early scores listed here. Obviously, the older you go, the better the gains are going to be. For what it's worth, comparing the iPhone 17 to the iPhone 13 shows a 71% gain in single-core and 79% gain in multi-core—not quite the 150% improvement Apple advertises. Perhaps there's something missing from the early scores here, or maybe Apple is using a different internal metric. We'll just have to see when testers benchmark the released iPhones running the public build of iOS 26.