Google’s long held a prominent role in the artificial intelligence field as a key source of research. Atransformer, one of the key “deep learning architectures” that contributed to the development of generative Al, was created by Google researchers. The “T” inChatGPTstands for “transformer.”

It’s not surprising then that the company was able to make generative AI a major focus of its products quickly. As is often the case when giant companies make quick pivots, Google hasn’t always done a perfect job naming its new products or explaining what they do. If you’ve heard ofGeminiandGemini Liveand know they have something to do with Google AI, but aren’t aware of what the relationship between the two is, here’s a way to understand the difference.

Google Gemini on Pixel 9 Pro

These 4 Google Gemini features make my life so much easier

Since switching to the Google Pixel 9 Pro, I’ve been relying on Gemini AI to take care of several previously mundane tasks.

The differences between Gemini and Gemini Live

Gemini is the model everything runs on

Gemini is the name of Google’s Al assistant available on the company’s website and directly from Android phones and iOS apps, but it’s also the multimodal, generative AI model that powers all of those things. You might already be able to see how things can get confusing.

Google introducedGemini 1.0as being available in three different sizes, depending on what you need the model to do and where it’s meant to run.

The Gemini website being used to generate images of animals.

Because Gemini is “multimodal,” it’s able to “understand” a variety of different inputs, not just text, but also images, files (if you have a code or a PDF), and audio. The Gemini chat interface can also produce a variety of different outputs thanks to integration with Google’s services and more specialized models, like Imagen 3, which can generate images. Gemini is the backbone that all of Google’s new AI models are built on, at least until the company introduces something even bigger and more capable, which is usually the goal with AI model development.

Gemini Live is a voice-based application of the model

In contrast, Gemini Live is the voice assistant feature integrated into the Gemini apps that lets you speak naturally to the AI model with your voice rather than text. If you first purchased a Google Home or Nest Audio device expecting to be able to have a human conversation with an AI, Gemini Live is a much closer realization of that idea. It’s also a direct competitor tothe voice modethat’s a part ofChatGPT, which also lets you talk to the model with natural language. The big difference is that Google hopes to integrate Gemini Live with apps like Gmail and Google Docs at some point in the future.

Even if its name suggests that it exists in parallel to Gemini, the fact that you access it in the Gemini app and it relies on the model itself, suggests it’s really more of a specific implementation of Google’s technology rather than a full distinct product.

A hand holding an iPhone 16 running Gemini Live in the Gemini app.

Unlike Google Assistant, talking to Gemini Live isn’t a collection of commands and responses. It’s conversational, and it’s designed to be interrupted and redirected when needed. Even if its name suggests that it exists in parallel to Gemini, the fact that you access it in the Gemini app and it relies on the model itself, suggests it’s really more of a specific implementation of Google’s technology rather than a full distinct product.

Gemini is the backbone of Google’s next big projects

Don’t be surprised if you see the name “Gemini” more often

Google has publicly referred to this current period of its product development as the"Gemini era,“as a testament to how serious the company is about AI going forward. Take it at its word or don’t, but using the Gemini model has already led to interesting experiences like Gemini Live. Now, you should hopefully have a better sense of the relationship between the two.

Google Gemini

Gemini is Google’s premier AI assistant app for the Android operating system that can provide text responses to questions, generate and analyze images, and is now available on iOS.

For more suggestions on how to use Gemini, Pocket-lint has suggestions on how to use the tool toselect the next book you reador use the AI model to bemore productive at work. Gemini is everywhere. It doesn’t need to be used everywhere, but there might be more than a few ways it’s useful to you.

Google Gemini icon