5 surprising Gemini 2.0 upgrades that actually improve Google's AI

Summary

Gemini 2.0 Flash generates images, text, and audio.
It can also autonomously complete tasks for you.
It’s even a virtual gaming assistant for real-time gameplay analysis.

Google’s newest AI update — Gemini 2.0 — is here, and it aims to be smarter, faster, and more capable than any other generative AI you’ve tinkered with before.

Unlike its predecessor, Gemini 1.5, which launched ten months ago, Gemini 2.0 can create images, produce audio with customizable voices, and even take actions on your behalf — like filling out forms or clicking links on a webpage. It can also run Google searches and even help you game better. In other words, it doesn’t just chat. The best part is it can handle all these new tasks while reportedly being twice as fast as the previous version.

Gemini 2.0 is actually a family of AI models, but the first one available is called Gemini 2.0 Flash.

Gemini 2.0 is actually a family of AI models, but the first one available is called Gemini 2.0 Flash. Developers can access it through the Google API in Google AI Studio and Vertex AI. But, if you’re not a developer, you can try it through the Gemini AI assistant on desktop or mobile. Just select the 2.0 Flash model from the app’s menu.

Advanced features like text-to-speech (TTS) and image generation are currently limited to Google’s early-access partners.

Gemini vs Apple Intelligence: What AI platform is better?

Gemini and Apple Intelligence are likely available on a device you already own, but which one is actually worth using?

1 Multimodal output

Gemini 2.0 Flash does it all

Gemini 2.0 Flash can now generate images, text, and audio as output, not just process them as input. It can create custom images from scratch with text, and it produces audio in multiple languages. You can even fully customize the voice of the audio. This is made possible by a new steerable text-to-speech (TTS) feature, which lets you control how the audio sounds. You can adjust the tone, speed, gender, and style, allowing it to sound formal, friendly, or casual, depending on your needs.

Google launches Gemini 2.0, ushering in the era of AI agents

The tech giant has launched Gemini 2.0 globally and revealed several of the experimental projects powered by it.

2 Agentic AI

Gemini 2.0 can take action for you

Agentic AI refers to AI bots that can autonomously complete tasks on your behalf, and Gemini 2.0 takes full advantage of this capability. Instead of just answering your questions, Gemini 2.0 can act for you. For instance, Google is unveiling Project Mariner, an experimental Chrome extension that can interact with web pages. It can click links, fill out forms, and navigate pages, making it a valuable tool for people who need help navigating unfamiliar websites or handling repetitive online tasks.

Google launches Gemini 2.0, ushering in the era of AI agents

The tech giant has launched Gemini 2.0 globally and revealed several of the experimental projects powered by it.

Agentic AI refers to AI bots that can autonomously complete tasks on your behalf.

3 Virtual gaming companion

Gemini 2.0 can observe your gameplay

Gemini 2.0 is powering a new gaming assistant that can observe gameplay in real-time and offer strategy tips to players. This builds on Google DeepMind’s history of using games to train AI in logic, planning, and rule-following. Now, by leveraging Gemini 2.0’s reasoning capabilities, Google’s assistant can analyze on-screen gameplay and suggest the next best move in real time.

Google is partnering with game developers like Supercell to test its ability to handle challenges in games like Clash of Clans.

A hand holding a Pixel 9 showing Gemini Live next to a laptop showing Gemini.

What’s the difference between Gemini and Gemini Live?

Google’s AI assistant already comes in a confusing number of flavors. Here’s the key difference between two of the more important versions.

4 Project Astra update

Now powered by Gemini 2.0

Project Astra now uses Gemini 2.0. Astra is a universal AI assistant that can recognize objects, help with navigation, and even locate misplaced items — like your glasses. With Gemini 2.0, Astra has the ability to converse in multiple languages, understand accents, and handle mixed-language conversations. It can also access tools like Google Search, Lens, and Maps, making it more useful for everyday assistance.

You can try out Project Astra by using the Gemini AI assistant on Android phones.

Google’s new Project Astra AI assistant amazed me, but it’s far from finished

Google’s latest digital assistant needs more work before it’s ready for consumer use.

Built into the Gemini app

Gemini 2.0 now directly leverages Google tools like Google Search, Google Lens, and Google Maps. Instead of having to open multiple apps, you can ask Gemini to handle it all in one place. Need to identify an object in a photo? Gemini’s got it. Want to search for a location and get directions? It’s all possible without switching apps.

Gemini 2.0 isn’t just another chatbot. It’s an actual assistant that can act, create, and customize its responses to meet your needs. From creating custom images and audio to taking actions on your behalf, Gemini 2.0 is smarter, faster, and more useful than ever before.

However, advanced features like text-to-speech (TTS) and image generation are currently limited to Google’s early-access partners.