Here are 10 huge Google product launches at Google Cloud Next 2025 today, including a new TPU, AI Agent Development Kit, Workspace innovation and Google’s new Cloud Wide Area Network.
From its new Ironwood TPU accelerator and Cloud Wide Area Network to a slew of new AI agents and Workspace innovation, Google unleashed 10 new products at Google Cloud Next 2025 today.
“Just one year ago, we shared a vision for how AI can fundamentally transform organizations. Today, that vision is not just a possibility – it’s the vibrant reality we are collectively building,” said Google Cloud CEO Thomas Kurian in a blog post today.
“The opportunity presented by AI is unlike anything we’ve ever witnessed. It holds the power to improve lives, enhance productivity, and reimagine processes on a scale previously unimaginable,” he added. “Today, at Next 2025, we’re proud to announce significant new innovations across our entire portfolio.”
[Related: Google Cloud CEO Thomas Kurian On Agentic AI’s Starring Role]
AI agents, or agentic AI, is Google Cloud’s largest investment this year as seen at Google Cloud Next with several new security and coding agents being unveiled.
Google’s conference is taking place in Las Vegas this week from April 9 to April 11.
Here are the 10 biggest product launches from Google Cloud Next 2025 today including a new AI Agent Development Kit, improvements to Vertex AI and Agentspace, as well as Google making its Gemini models available on-premises.
New AI Agent Development Kit
Google Cloud added new capabilities to help customers move towards a multi-agent ecosystem with the launch of its new AI Agent Development Kit (ADK).
“Our new AI Agent Development Kit is an open-source framework that simplifies the process of building sophisticated multi-agent systems while maintaining precise control over agent behavior. With ADK, you can build an AI agent in under 100 lines of intuitive code,” said Kurian.
ADK supports Model Control Protocol to allow developers to easily plug in available tools. ADK also connects seamlessly with APIs managed by Apigee to leverage a user’s own business APIs.
As part of ADK, Google Cloud unveiled Agent Garden, which is a new collection of ready-to-use samples and tools directly accessible in ADK. “It allows you to connect your agents to 100-plus pre-built connectors, your custom APIs, integration workflows, or data stored within your cloud systems like BigQuery and AlloyDB,” Kurian said.
Google Cloud also said customers can now build AI agents directly on their existing NetApp data, no data duplication required.
Ironwood TPUs
One of the biggest launches at Google Cloud Next 2025 is Google’s 7th-generation TPU: Ironwood. The Ironwood AI accelerator was specifically designed to improve performance and scalability for inferencing.
This advancement will enable Google and its cloud customers to develop and deploy more sophisticated inferential AI models at scale, Kurian said.
“Our seventh generation TPU, Ironwood represents our largest and most powerful TPU to date, a more than 10x improvement from our most recent high-performance TPU,” said Kurian.
Featuring over 9,000 chips per pod, Ironwood delivers a staggering 42.5 exaflops of compute per pod, meeting the exponentially growing demands of the most sophisticated thinking models, like Gemini 2.5, Kurian added.
Google’s New Cloud Wide Area Network
Google’s massive global network infrastructure includes 42 regions connected by more than 2 million miles of terrestrial and subsea cables, with more than 200 points of presence across over 200 countries.
“Starting today, this network, which moves at ‘Google speed’ –near-zero latency—for billions of users worldwide, is now available to enterprises everywhere. We call it Cloud Wide Area Network (or Cloud WAN),” said Kurian.
The new Cloud WAN makes Google’s global private network available to all Google Cloud customers.
“Cloud WAN is a fully managed, reliable, and secure enterprise backbone to transform enterprise wide area network architectures. It delivers a remarkable improvement of up to 40 percent in network performance, while simultaneously reducing total cost of ownership by up to 40 percent,” said Kurian.
Cloud WAN enables two key use cases: providing high-performance connectivity between geographically dispersed data centers, and connecting branch and campus environments. At launch, Cloud WAN partners include Accenture, HCLTech, and Wipro, who now offer services architecture, design, migration and ongoing management services.
Gemini On-Premise Via Google Distributed Cloud
Google announced that Gemini will be available on Google Distributed Cloud (GDC), bringing Google’s most capable models to on-premises environments, with public preview starting in the third quarter of 2025.
“We have partnered with Nvidia to bring Gemini to Nvidia Blackwell systems, with Dell as a key partner, so it can be used locally in air-gapped and connected environments. This compliments our GDC air-gapped product, which is now authorized for U.S. Government Secret and Top Secret levels, and on which Gemini is available, provides the highest levels of security and compliance,” said Kurian.
GDC is a fully managed on-prem and edge cloud solution that is offered in both connected and air-gapped options, scaling from a single server to hundreds of racks. It offers Infrastructure-as-a-service, security, data, and AI services.
Google Workspace New Innovation
Google unleashed a slew of new AI innovations for Workspace, which includes its popular applications like Gmail, Docs, Drive and Meet.
Help me Analyze is a new feature that transforms Google Sheets into a user’s personal business analyst, intelligently identifying insights from data without the need for explicit prompting.
Workspace Docs’ new Audio Overview lets users create high-quality, human-like audio read-outs or podcast-style summaries of your documents.
The new Google Workspace Flows helps users automate daily work and repetitive tasks like managing approvals, researching customers, organizing emails and summarizing daily agendas.
Vertex AI’s Four New Advancements
Google’s Vertex AI platform enables the building and managing of AI applications and agents, as well as model training and deployment. Vertex AI usage experienced a 20X increase in growth last year, resulting in thousands of AI applications built by customers.
Google unveiled four new key enhancements to Vertex AI at Google Next, including new Dashboards to help monitor usage, throughput, latency, and troubleshoot errors.
Another improvement came via new training and tuning capabilities. Google said users can now manage custom training and tuning with their own data on top of foundational models in a secure manner across all first party model families including Gemini, Imagen, Veo, embedding, and translation models, as well as open models like Gemma, Llama, and Mistral.
Vertex AI’s new Model Optimizer leverages Gemini to automatically direct users query to the most performant model and tools, based on their quality, speed and cost preferences.
The fourth new enhancement to Vertex AI is dubbed Live API. To enable conversational interactions, Live API offers streaming audio and video directly into Gemini to let agents process and respond to media in real time.
Agentspace 5 New Capabilities
Agentspace brings together Google’s enterprise search, conversational AI, Gemini and third-party agents to enable employees to find and synthesize information from within their organization, converse with AI agents, and take action with their enterprise applications. It delivers a set of tools including pre-built connectors to search and transact with documents, databases, and SaaS applications, as well as advanced security and compliance to protect data.
At Google Cloud Next, the company unveiled five new enhancements to Agentspace including new seamlessly integration with Chrome Enterprise, allowing employees to search and access all their enterprise resources directly from the search box in Chrome.
Another new capability unveiled today is Agent Gallery, which provides employees a single view of available agents across the enterprise—including those from Google, internal teams, and partners to make agents easy to discover and use.
The third new innovation on Agentspace is dubbed Agent Designer. This is a no-code interface for creating custom agents that automate everyday work tasks or enhance knowledge, helping employees adapt agents to their individual workflows and needs.
The fourth new enhancement is Google Cloud’s new Idea Generation agent. This agent uses a tournament-style framework to effectively rank ideas based on employee-defined criteria, and can help employees refine or generate new ideas, brainstorming, and problem-solving.
Another new agent from Google is its Deep Research agent, which explores complex topics on a user’s behalf to provide findings in a comprehensive, easy-to-read report.
New Security Innovation
At Next, Google unveiled its new Google Unified Security solution which brings together visibility, threat detection, AI-powered security operations, continuous virtual red-teaming, the most trusted enterprise browser, and Mandiant expertise—in one converged security solution running on a planet-scale data fabric.
It creates a single, scalable and searchable security data fabric across the entire attack surface. The solution provides visibility, detection and response capabilities across networks, endpoints, clouds and applications. It automatically enriches security data with the latest Google Threat intelligence for more effective detection and prioritization.
Another security innovation launched at Next, is Google’s new security agent, dubbed Alert Triage Agent, performs dynamic investigations on behalf of users. “It analyzes the context of each alert, gathers relevant information, and renders a verdict on the alert, along with a history of the agent’s evidence and decision making,” said Kurian.
Google also unveiled its new Malware Analysis Agent that investigates whether code is safe or harmful. “It builds on Code Insight to analyze potentially malicious code, including the ability to create and execute scripts for deobfuscation,” said Kurian.
Gemini 2.5 Flash For Vertex AI
Last month, Google launched Gemini 2.5 Pro for Vertex AI, optimized for precision and for writing and debugging intricate code or extracting critical information in medical documents.
“Today we’re announcing Gemini 2.5 Flash—our workhorse model optimized specifically for low latency and cost efficiency—is coming to Vertex AI,” said Kurian.
Gemini 2.5 Flash adjusts the depth of reasoning based on the complexity of prompts, and users can control performance based on customers’ budgets.
“Flash is ideal for everyday use cases like providing fast responses during high-volume customer interactions, where real-time summaries or quick access to documents are needed,” said Kurian. “These new features make powerful AI easier to use and more affordable for everyday use cases, enabling our customers to build AI that solves complex problems and understands nuance.”
New Coding Agents Innovation
Google’s Gemini Code Assist for Google Cloud, Android Studio and Firebase Studio lets customers understand their company’s code base, standards, and conventions.
At Next today, Google unveiled new Code Assist agents to help with everything from modernizing code to assisting with the full software development lifecycle.
New offerings include a kanban board, which lets developers interact with Google’s agents on a kanban board.
Code Assist also now has integrations from partners such as Atlassian, Sentry and Snyk. Outside of Google, Gemini is now also available for customers development needs in Aider, Cursor, GitHub Copilot, Replit, Tabnine, and Windsurf.
