Google Cloud used its Gemini at Work 2026 event to announce the "Gemini agent," described as a new single, universal agent for work. According to 9to5Google, the agent is available in the Gemini Enterprise app, Google Workspace and third-party services, and it is designed so that users give it objectives rather than step-by-step instructions.
The pitch is a shift from simple assistants to something closer to a colleague. Google says the Gemini agent can answer questions, handle knowledge work, create media, and write and run code, and it can act either as a personal assistant or as a team member working on behalf of a group or a specific role, such as an analyst in a finance department.
Gemini agent features and multi-model support
A key detail is that the Gemini agent is not locked to Google's own models. It runs each job on the model that fits best, which today includes both the Gemini and Claude families, with support for other private and open models promised in the future. Google frames this as a way to deliver optimal quality while lowering costs.
The agent uses sub-agents to tackle multi-step tasks and connects with third-party tools, coordinating workflows that can run in parallel or in sequence for hours or even days. Because it runs in the cloud, it keeps a single set of memories, context and one personalization graph, with four kinds of memory. In Gmail, Google Docs, Sheets, Slides and Chat, users can mention @Gemini to invoke it.
- Availability: Gemini Enterprise app, Google Workspace and third-party services
- Channels: command line, Google Workspace, Slack and Microsoft 365, from any device



