The company on Wednesday announced its latest phone models, the Pixel 8 and Pixel 8 Pro. Among a swath of new features—including brighter screens and upgraded photo editing capabilities—the phones lean heavily into artificial intelligence technology. 

“It’s the first phone engineered and built for the generative AI era,” Rick Osterloh, senior vice president of devices and services, said of the more advanced Pro model during the launch event. Included in the phone’s new features, it can generate a background fill for photos and summarize voice notes. 

While Google’s Android software is the leading smartphone operating system, Google’s phone hardware business has room to grow. The Pixel line of devices has less than a 1% global market share, Reuters reported, with Apple and Samsung the industry leaders. It isn’t clear how much money Google makes off its phones because it clumps together hardware products in earnings reports, but its phone line is prepped to grow in relevance. The Pixels it launched last year were the best-selling generation in Google’s history, CEO Sundar Pichai said during the company’s February earnings call.

As Google looks to build on that momentum, it’s turning to an asset that Apple and Samsung can’t match: a massive generative AI business. On the back of Google’s Bard launch in March, generative AI is quickly becoming a key technology at the company, threading through its cloud, workspace and hardware products. Apple is said to be building out a large language model called Ajax to rival Bard and ChatGPT, according to Bloomberg. But it has thus far released nothing of the sort for commercial use, nearly one year after ChatGPT’s launch sparked a development frenzy in the industry. Apple’s consumer-facing AI technologies have largely focused on improving autocorrect and dictation in phones. Samsung is similarly using AI to improve its autocorrect, dictation and phone cameras, but other applications of the tech are unclear. 

Google’s Osterloh said Wednesday that the company was adding a “distilled” versions of its text- and image-generating models to the Pixel, a feature he claimed makes Pixel the first smartphone to run the company’s generative AI independently on a device.

Using the new Pixel 8 Pro model, photographers can drag objects out of a picture, and AI will generate the background to fill the space. It can unblur objects and sharpen minute details in a photo as well. For its audio capabilities, Pixel can provide real-time transcriptions and summarizations for voice recordings. A virtual assistant can not only answer the phone for users, but hang up on spam callers. When multilingual users dictate to the phone, AI can switch between languages as it types—something Monica Gupta, head of strategy and operations for data management, said no other phone can do. 

While some of these features come with the Pixel at launch, others will arrive with a software update in December, the company said. At Wednesday’s event, the company also teased “Assistant with Bard” on Gmail and Google Drive applications, which can summarize emails and view users’ screens to help with tasks. This feature could find its way onto non-Google phones, given that rival app stores offer these Google apps, but details on the roll-out are unclear.