Gemini Enterprise (cloud.google.com)

Google Cloud today launched Gemini Enterprise, a unified, enterprise-grade platform that brings the Gemini family of models, purpose-built TPUs, and DeepMind research into a single “front door” for workplace AI. Gemini Enterprise combines six core pieces — advanced Gemini models, a no-code workbench, prebuilt and custom agents, secure connectors to data across Google Workspace, Microsoft 365 and enterprise apps, centralized governance, and an open partner ecosystem — to move beyond point AI tools toward automating end-to-end workflows. Google also highlighted multi-modal agents in Workspace (video generation with Google Vids, Meet real-time speech translation), a Data Science Agent preview for automated data wrangling and model pipelines, and next-gen conversational agents with a low-code builder supporting 40+ languages. Technically this bundles model, infra and orchestration: Vertex AI integration, Gemini models powering agents and CLI, and extensible Gemini CLI extensions for developer toolchains. Google is pushing standards for an “agent economy” with open protocols — Agent2Agent (A2A), Model Context Protocol (MCP) and a new Agent Payments Protocol (AP2) with major payment partners — enabling agent-to-agent communication and secure transactions. Early customers (Banco BV, Harvey, Commerzbank, Klarna, Mercedes, Figma) show use cases from automated analytics and legal work to conversational CX and personalized commerce. The result: faster time-to-value, centralized governance for security/compliance, and a platform that aims to replace stitched-together AI stacks with an integrated, production-ready fabric for enterprise AI.