Gemini 2.5, supercomputing, intelligent agents, and no-code tools mark Google’s most ambitious cloud event to date.
Google has pulled out all the stops at its Cloud Next ’25 event held this week, where it unveiled over 3,000 technological advancements aimed at solidifying its leadership in artificial intelligence, cloud infrastructure, and business productivity. Far from being a theoretical exhibition, the company demonstrated how its ecosystem—driven by models like Gemini, platforms like Vertex AI, and tools like Agentspace—is already transforming organizations worldwide.
Gemini 2.5: AI for thinking, reasoning, and decision-making
The star of the event was Gemini 2.5, Google’s most advanced family of models. Specifically, two versions tailored for different uses have been launched:
- Gemini 2.5 Pro, designed for complex tasks like code generation or medical analysis.
- Gemini 2.5 Flash, built for ultra-fast and cost-efficient responses, ideal for customer service or intensive-use tasks.
Both models are now available on Vertex AI, where the usage of Gemini has increased 40-fold in the past year.
AI Hypercomputer: the supercomputer training the AI of the future
Google also unveiled its new system, AI Hypercomputer, a combination of hardware and software optimized for artificial intelligence workloads. Key announcements include:
- TPUs Ironwood, featuring over 9,000 chips per pod and up to 42.5 exaflops of computing power.
- New A4 and A4X GPUs powered by NVIDIA Blackwell technology.
- Ultra-low-latency storage systems like Anywhere Cache and Rapid Storage.
This environment allows for reduced costs, accelerated model training, and greater power without compromising energy efficiency or response time.
Agentspace: AI for every employee, no coding required
One of the most accessible proposals is Google Agentspace, which lets any employee create and use AI assistants without technical knowledge. It offers features like:
- Agent Designer, a no-code interface for designing custom agents.
- Deep Research Agent, which analyzes large volumes of information and provides summarized reports.
- Integration with the Chrome browser to access corporate resources from the search bar.
With clients like KPMG, Rubrik, and Wells Fargo already incorporating it, Google aims to democratize access to artificial intelligence within companies.
Workspace and productivity: AI integrated into everyday tasks
Google has also revamped its productivity suite, Google Workspace, which includes Gmail, Docs, and Sheets. The new AI features are clear bets for efficiency:
- Help Me Analyze in Sheets, to draw conclusions from data without formulas.
- Docs Audio Overview, which generates audio summaries of documents.
- Workspace Flows, a system for automating routine tasks like approvals or email summaries.
Currently, Workspace provides over 2 billion AI assists per month.
Models for images, voice, music, and video
The event also showcased Google’s creative muscle with advancements in generative models:
- Imagen 3: improved precision in image generation and editing.
- Veo 2: advanced video editing via AI.
- Lyria: music creation from text.
- Chirp 3: generation of personalized voices with just 10 seconds of audio.
These models are already being used by companies like Agoda, Kraft Heinz, and Spotify to create content at scale and optimize marketing campaigns.
Security, data, and a commitment to multicloud
In cybersecurity, Google launched Unified Security, a comprehensive platform that combines threat detection, malware analysis, and specialized agents like the Alert Triage Agent and the Malware Analysis Agent.
Additionally, it reinforced its multicloud strategy, with native integration with Amazon, Azure, SAP, Oracle, and Salesforce, and introduced advancements in its Google Distributed Cloud solution, which enables AI models to be deployed in local or regulated environments.
With Cloud Next ’25, Google shows that its commitment to artificial intelligence goes well beyond mere rhetoric. It is creating a cohesive, powerful, and, above all, useful ecosystem for companies across all sectors. AI is no longer a promise: it is a platform for real, scalable transformation, available today.
Source: Artificial Intelligence News