Google Cloud announces Gemini Agent, integrating Workspace and multi-model routing

AT A GLANCE

The report says Google Cloud announced Gemini Agent, integrating office applications, multi-model routing, persistent execution, and enterprise governance.

Article

Google Cloud announces Gemini Agent: a single general-purpose work agent integrating Workspace, multi-model routing, and enterprise governance

Beating AI news flash, Google Cloud CEO Thomas Kurian announced at Gemini at Work 2026 the conference Gemini Agent, positioning it as a single general-purpose work agent—users only need to provide goals rather than instructions, and it can complete knowledge work, content creation, image and media generation, and code writing and execution. The agent offers a unified entry point, ubiquitous access (web, iOS, Android, Windows, Mac, and Workspace、Microsoft 365, Slack, and other channels), persistent execution (running in the cloud so tasks continue after the laptop is closed, with memory and personalization profiles unified across devices), multi-agent orchestration (it can dynamically create temporary sub-agents with independent identities, or serve as a long-term colleague agent with its own email and storage), deep contextual understanding, and model-selection flexibility—the underlying model can be freely switched among Google’s Gemini series and Anthropic’s Claude, among others, to match the best model to each task and reduce costs. Currently, nearly 80%of Google Cloud customers use its AI product, and nearly 90%of Fortune 100 companies use Gemini Enterprise.

The agent is embedded directly in Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar, carrying the same memory, skills, and controls. A new data-analysis skill enables both technical and business users to obtain actionable insights in minutes using natural language; it also provides industry-specific tools and knowledge for financial-services and legal teams. For security and governance, Google provides mechanisms including identity and policy management, authorization and access controls, secure sandboxes, and network gateways. For cost control, it uses multi-model orchestration, intelligent routing, and real-time spending caps. Early test customers include sports brand On, which uses dynamic model selection to accelerate time to market; Shopify mixes frontier models for millions of merchants, while PayPal routes 10000000 multi-model requests per week.

Original link https://m.theblockbeats.info/flash/370829

Key points

Beating AI news flash reports that Google Cloud CEO Thomas Kurian announced at Gemini at Work 2026 the conference Gemini Agent。
The report positions Gemini Agent as a single general-purpose work agent that, after users provide a goal, can be used for knowledge work, content creation, image and media generation, and code writing and execution.
The access channels listed in the report include web, iOS, Android, Windows, Mac,Workspace、Microsoft 365 and Slack.
The report says the agent can continue running in the cloud and supports unified memory and personalization profiles across devices.
The report says the agent can dynamically create temporary sub-agents with independent identities, or serve as a long-term colleague agent with its own email and storage.
The report says the underlying model can switch among Google Gemini series and Anthropic’s Claude, among other models.
The report states that nearly 80%of Google Cloud customers use its AI product, and nearly 90%of Fortune 100 companies use Gemini Enterprise。
The report says the agent is embedded in Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar, and carries the same memory, skills, and controls.
The security and governance mechanisms listed in the report include identity and policy management, authorization and access controls, secure sandboxes, and network gateways; cost-control methods include multi-model orchestration, intelligent routing, and real-time spending caps.
The report says early test customers include On, Shopify, and PayPal, and states that PayPal routes 10000000 multi-model requests per week.

AI-assisted interpretation

The following is analysis, separate from reported facts. Verify important claims independently.

This news flash describes a unified work agent: after users state a goal, it can call office applications and different models to perform tasks and continue running in the cloud. The report also emphasizes that enterprises can manage identities, permissions, secure environments, and spending. All of the above are claims made in this news flash; the available materials do not include an official Google Cloud announcement for verification.

Why it matters to readers

If the reported functionality is provided as described, enterprises may be able to handle office, content, data-analysis, and code tasks through a single entry point, while switching models according to the task and controlling permissions and costs.

For beginners,Gemini Agent can be understood as a work assistant that connects to various office tools and models; however, the current materials do not explain whether it can actually be used, how it is priced, or what functional limitations apply.

Risks and unknowns

The only existing evidence is a media news brief; no official Google Cloud announcement or independent verification materials were provided.
The product’s official launch date, available regions, and pricing are not specified in the evidence.
The specific support scope, limitations, and partnership status of the third-party channels and the Claude models listed in the report cannot be confirmed.
Customer usage proportion, wealth 100 strong adoption rate, and PayPal request volume all lack original statistical definitions.

Google Cloud announces Gemini Agent, integrating Workspace and multi-model routing

Article

AI-assisted interpretation

Why it matters to readers

Risks and unknowns

Related Developments

Related concepts

Keep following

Security Center