Hey everybody, it’s new phone time!

Get ready for CEOs and product leaders to take the stage and give a flashy sales pitch for sleeker designs, better cameras, and smarter functionality. As of now, Google has made their product announcement, and Apple is preparing theirs for September 9th.

Yet these releases, we’ve been thinking about how Google and Apple are implementing large language models (LLMs) within their devices. Whether you’re already embracing AI on your phone or are curious about how to do it, we have some insights for you.

**We understand that the iOS release is imminent and will update our observations after the launch.

Android

Most of us are familiar with Gemini, Google’s homegrown large language model. In recent months, Google has been rolling out a wave of new features and improvements aimed squarely at making Gemini more capable, more connected, and more useful in day-to-day workflows.

With the Pixel 10, Google’s devices now have access to the Tensor G5. This powerful silicon chip is built to work with Gemini Nano, their lighter AI model, which can run without an internet connection and is tailored to enhance many common mobile tasks. Most importantly, it’s designed to integrate with your entire Google ecosystem, offering some key features that make it a standout assistant:

  • Gemini Live – Debuted earlier this year, this feature allows you to ask Gemini about anything on your screen or that your camera lens can see. Google expanded the capabilities of Gemini Live to interact with other apps in their ecosystem and introduced on-screen guidance to provide smart suggestions and information about what your camera sees. That means their AI chatbot not only can extract information from those apps but can pivot on a dime to address your changing questions and real-time thoughts.

  • Gemini Advanced – The premium version of Google’s AI assistant, Gemini Advanced builds on the standard experience with enhanced reasoning, deeper integration, and priority access to Deep Research. It can manage complex, multi-step requests (tasks like researching a market trend, drafting content, and summarizing findings) while drawing on data across Gmail, Drive, and other Google apps. Designed to be more proactive, it can anticipate follow-up questions, refine outputs in real time, and help users navigate large projects from start to finish without losing context.

  • Magic Cue – Google’s stated goal for their phones is to “anticipate your needs and make your life easier.” With Magic Cue, they’re taking a major step forward. This AI-powered feature reviews voice and text conversations in real-time to surface relevant links and details when you need them. This saves time and makes interactions more natural, which you can’t beat when you’re trying to communicate.

  • Imagen 4 – Focused on creating a cohesive experience across their AI capabilities, Google has integrated their Imagen 4 text-to-image model with the Gemini API. This allows users to access the sharper and more detailed images of the latest version of Imagen, though there are still some issues with saturation, blur, and editing.

Now, that doesn’t mean it’s fully up-to-speed or without flaws. There are some tasks that Google Assistant was able to accomplish with greater dexterity that Gemini struggles to manage. Up until a few months ago, adding a new to-do to the Task app or turning off one of your smart home devices wasn’t possible with Gemini. Yet Google has been making rapid strides to elevate their AI performance and reestablish any lost functionality.

On a more bizarre note, some users are reporting that the AI chatbot doesn’t handle failure well, wallowing in melancholy histrionics that have the Google Gemini team baffled. So, you might need to talk your chatbot off the ledge after your next convo, but the tech giant is trying to get Gemini some therapy.

iOS

Apple Intelligence is the AI system Tim Cook and team released as the answer to ChatGPT. It’s their umbrella term for AI features across their iOS devices. You’ve likely seen some of the commercials in their media blitz, quirky teasers of their promised suite of features, but the rollout has been slower than promised.

The iPhone 16 was built to handle Apple Intelligence but lacked announced AI features at launch. In subsequent updates, iOS devices have been able to access a number of budding tools:

  • Writing Tools – Apple has woven AI-assisted writing directly into Mail, Notes, and third-party apps. Users can rewrite text for clarity, adjust tone, or generate new content from a short prompt, all within the app. Plus, the integration with ChatGPT can help to elevate the quality of text generation.

  • Notification Summaries –The feature uses on-device intelligence to group and condense incoming notifications into quick, relevant digests. After a rocky start, Apple Intelligence is once again allowing summaries for news and entertainment apps as well as the others using their operating system.

  • Visual Intelligence – Similar to Google Lens, this tool lets you identify objects, landmarks, documents, or even handwritten text directly from your photos or camera view. It can also act on that information by researching details, adding events to your calendar, or initiating relevant tasks.

Then, there’s the ChatGPT integration. iOS 26 will get an upgrade to the newest GPT-5 model and will be able to respond to quick questions and delve into deeper analysis as well. However, there’s a slight caveat: You need to activate the chatbot within the iOS before you can access its functionality:

Go to Apple Intelligence & Siri->Turn on Apple Intelligence->Scroll down to Extensions->Click on ChatGPT->Setup ChatGPT->Review the pop up telling you what you can do->Enable ChatGPT or Use ChatGPT with an Account

Not a difficult set of steps, but it’s still extra effort to run on your operating system. Moreover, Siri itself doesn’t have the full capabilities of ChatGPT or Gemini, so if you stump it, the proprietary virtual assistant will send your request to the third-party app.

Though Tim Cook, Apple CEO, recently declared they would be prioritizing AI investment, they’re still short of the benchmark, and the scuttlebutt is that we won’t hear about any Siri updates at WWDC 2025.

Keeping Eye on the News

What’s important to know is this can change quickly. As Tim Cook reportedly said in his all-hands meeting, “We’ve rarely been first. There was a PC before the Mac; there was a smartphone before the iPhone; there were many tablets before the iPad; there was an MP3 player before iPod.” He’s pointing to what Marque Brownlee has called their “second-mover advantage,” and they could throw a Hail Mary with an unheralded Siri update that wins the AI game.

However, that might be outside of their wheelhouse. All Apple’s most revolutionary contributions have been physical products. iMac. iPhone. iPad. All gamechangers, all hardware. 

On top of that, they’ve been unafraid to outsource hardware components or software. The Apple iPhone 16, which has one of the better OLED displays of U.S. smartphones, uses digital displays manufactured by Samsung Display. Even Siri, which is synonymous with the brand, was developed by a third-party company, which Apple acquired and has since maintained.

So, even though Android and iOS are adding more LLM-powered mobile AI tools, they’re moving at very different speeds and with different priorities:

  • Google has leaned into rapid feature expansion, betting on ecosystem-wide integration to keep Gemini ahead of the curve.
  • Apple, meanwhile, is working behind-the-scenes to refine their features before pushing them broadly (hopefully learning from some mistakes).

For business and everyday users alike, the real story is how quickly this landscape is evolving. A year from now, the balance could look very different, and whichever device you carry, it’s worth keeping an eye on how these AI tools mature and (just as importantly) how seamlessly they fit into the way you already work and live.

Want to learn more about the latest technology? Check out our Cubex Group blog for our latest thoughts, tips, and tactics.

Related Articles

AI Personal Assistants for Business: Picking the Right Fit for Your Operations

Why You Need a Business VPN Strategy in 2025

Why One New iOS 18 Feature Is a Blessing to Communication AND a Threat