Google Cloud has announced the Gemini agent, which it describes as a universal agent for work.
The company introduced the product at its Gemini at Work 2026 event.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
The agent handles knowledge work, question answering, content creation and coding from a single prompt box.
According to Google Cloud, it plans tasks, uses skills and tools, and connects to customers’ business systems. It returns finished work inside the documents, inboxes and developer environments that people already use.
The company said it also selects a model for each job, includes built-in cost controls, and provides the security, administration and governance that enterprise customers require.
Google Cloud said users give the agent objectives rather than instructions. It can work as a personal assistant or as a team member acting for a group, such as a project manager, or for a role, such as an analyst in a finance department.
The agent can also create images and media, and write and run code, through a single agent and a single API. Users can assign work, schedule tasks or have the agent respond to events.
Users can reach the agent from the web, iOS and Android devices, and Windows and Mac desktops. Channels include the command line, Google Workspace, Microsoft 365 and Slack. It can also be built into third-party applications and can operate as a headless agent, meaning it needs no dedicated user interface.
Google Cloud said the agent runs in the cloud and keeps one set of memories, context and a single personalisation graph across devices and channels. Work that takes hours or days continues after a user closes a laptop.
The agent can create sub-agents, which are temporary, job-specific agents with their own identities. It coordinates them across parallel and sequential steps that can run for hours or days. It can also act as a coworker agent, with a persistent, defined role that spans days, sessions and changing responsibilities.
On models, Google Cloud said the agent and the model beneath it are separate choices. Currently, the agent orchestrates across the Gemini family of models and Claude models from Anthropic. The company said it plans to add other private and open models in the future.
