Google Cloud is consolidating its enterprise artificial intelligence (AI) stack into a single system designed to build, deploy, and manage autonomous software agents at scale, as companies move beyond simple AI prompts toward full task automation across business systems.
The new Gemini Enterprise Agent Platform expands on Vertex AI and brings model development, agent orchestration, security controls, and deployment tools into one environment. It is designed as a central hub where enterprises can design AI agents that can operate across tools, workflows, and internal systems with less manual oversight.
The platform gives access to more than 200 models through Model Garden, including Google’s Gemini 3.1 Pro, Gemini 3.1 Flash Image, and Lyria 3, along with open models such as Gemma 4 and third-party models like Anthropic’s Claude Opus, Sonnet, and Haiku.
Google Cloud said all Vertex AI services and future updates will now be delivered through the Agent Platform, effectively making it the company’s main foundation for enterprise AI agent development moving forward.
The platform is designed to shift enterprises away from managing individual AI tasks and toward delegating full business outcomes to agents.
Under the Build layer, users can develop agents through Agent Studio, a low-code visual interface, or the Agent Development Kit (ADK) for code-based development. The company also added AI-native coding support to speed up production-ready builds.
For deployment, the upgraded Agent Runtime supports long-running agents that can maintain state for days, backed by Memory Bank for persistent context retention across tasks.
On governance, the platform introduces Agent Identity, Agent Registry, and Agent Gateway, which assign each agent a trackable identity and enforce enterprise security and access controls across systems and partner-built agents.
To improve reliability, Agent Simulation, Agent Evaluation, and Agent Observability provide execution tracing and real-time insight into how agents make decisions and perform tasks.
The shift shows that rising demand for AI systems can do more than generate responses, particularly in industries where automation needs to span multiple tools, workflows, and data systems.
By combining model access, agent orchestration, governance, and monitoring in one platform, Google Cloud is promoting Gemini Enterprise Agent Platform as an infrastructure layer for enterprise AI adoption rather than a standalone toolset.
The inclusion of more than 200 models, including proprietary and third-party systems, also signals a multi-model strategy, allowing businesses to choose models based on cost, performance, and task complexity.