When we gathered at Google Cloud Next a year ago, we talked about how generative AI was beginning to change the way we work. This week in Vegas, we showcased how AI is not only transforming work, it’s running at scale.

We’ve officially entered the agentic era. AI is becoming an active partner — an agent that can actually do the work for you safely and autonomously. At Next ‘26, we shared everything you need to build your own helpful, intelligent agents and use them to grow your business. Here are the highlights.

1. The new Gemini Enterprise Agent Platform is here.

For the technical teams building the future, we’re introducing the Gemini Enterprise Agent Platform. It’s a complete, end-to-end workspace to build, govern and scale your AI agents with the world’s best models at your fingertips.

We’re providing direct access to Gemini 3.1 Pro, our most capable model yet for handling complex workflows, alongside Gemini 3.1 Flash Image (also known as Nano Banana 2) for creating stunning visual assets, and Lyria 3 for professional-grade audio. We’re also expanding our commitment to open choice by adding Anthropic’s Claude Opus 4.7.

You don’t have to be a machine-learning PhD to use these models or build an agent. Agent Studio, our low-code interface, lets developers and business users alike build and test agents using natural language.

2. The Gemini Enterprise app brings AI to your everyday work.

We’re bringing the power of the agents directly to your workforce through the Gemini Enterprise app. To ensure everyone can tap into this technology, our no-code Agent Designer lets anyone build custom, trigger-based workflows without writing a single line of code.

For more complex, multi-step business processes, new long-running agents can work autonomously in the background within secure cloud sandboxes while you focus on other things. With all these agents helping out, we’ve built a central, intuitive Agent Inbox so you can easily monitor, guide and manage exactly what your AI helpers are up to.

3. Our new eighth-generation TPUs are designed to power the AI era.

Running millions of AI agents takes some serious computing muscle. Our AI Hypercomputer is a purpose-built system designed specifically for the massive scale of this new era, including the eighth generation of our custom TPU (Tensor Processing Unit) chips.

The TPU 8t is built to train AI models incredibly fast, while the TPU 8i is optimized for inference (actually serving up the models), delivering 80% better performance per dollar. We will also be among the first to offer the new NVIDIA Vera Rubin NVL72 systems, joining our existing lineup of NVIDIA GPUs and our super-efficient Google Cloud Axion processors.

To take advantage of this powerful compute, you need to move data at lightning speed. We unveiled the Virgo Network, a custom-built system to connect massive supercomputers, alongside storage breakthroughs like Managed Lustre, which can now move an incredible 10 terabytes of data per second.

4. The new Agentic Data Cloud helps make sense of your data.

An AI agent is only as helpful as the information it understands. We’re introducing the Agentic Data Cloud as a completely new way of organizing your data so AI can take action in real time. The Knowledge Catalog acts like a dynamic map of your entire business. Using Gemini, it autonomously tags and connects the dots across your enterprise so your agents actually understand the unique context and lingo of your company.

The Agentic Data Cloud also includes Cross-Cloud Lakehouse, because you shouldn’t have to move all your data to use our AI. Standardized on Apache Iceberg, this lets you leave your data exactly where it is — even if it’s in AWS — and query it instantly, with zero friction.

5. We’re building security for the AI age.

As cybersecurity threats get smarter, human defenders need AI on their side. We’re combining Google’s world-class threat intelligence with Wiz’s leading security platform to give you agentic defense. Specialized new agents include the Threat Hunting agent that can proactively hunt for threats and write security rules autonomously, the Detection Engineering agent that can identify coverage gaps and create new detections, and the Third-Party Context agent that provides contextual data from third-party content.

Wiz, now a part of Google Cloud, expands and deepens our ability to protect the apps you build and run. Wiz has added Databricks to the list of cloud platforms it supports, as well as more AI studios and multicloud platforms-as-a-service.

Additionally, the new Technology Intel Center aggregates and centralizes a feed of new features releases, migration updates, and end-of-life notices across cloud and AI tech providers that are relevant to you.

6. Workspace Intelligence will help teams collaborate even better.

We’re baking agentic power right into the tools you use to connect with your team and your customers. With Workspace Intelligence, we’re breaking down the walls between Docs, Drive, Meet and Gmail. You can use Ask Gemini in Chat to synthesize info from across your Google Workspace and take immediate action — like scheduling a meeting or drafting a brief — without ever leaving your chat window.

7. Companies are now full of AI builders.

The best part of Next is seeing what people are already doing with this technology.

The Home Depot is using Gemini to power a phone and in-store assistant that gives shoppers expert advice and finds answers faster. Papa John’s is using our Ordering Agent, which uses AI context to remember “the usual” order and get diners what they want faster.

Mars and Citadel Securities are using AI tools to streamline their quantitative research, whether that’s for marketing or market making. And Unilever is deploying agents throughout its organization to serve its 3.7 billion consumers.

https://blog.google/innovation-and-ai/infrastructure-and-cloud/google-cloud/google-cloud-next-26-recap/