AGENT PROFILES

Gemini agent

Google · General-purpose agents

Universal work agent announced on October 8, 2026 for research, content and code. Google describes shared enterprise context, tools and persistent cloud execution.

Gemini · Announced universal cloud agent

Product and access

Gemini agent combines research, content and code in an enterprise context. The Gemini product family also includes Gemini CLI for terminal tasks and Gemini Spark for personal tasks.

Provider

Google Cloud develops the Gemini agent for enterprise work. It connects Gemini models with Workspace, enterprise data and administratively managed tools.

Operator
Google
Organization
Technology company
Provider sourcesSource 1 Source 2 Reviewed
01

Core features

9 features from provider sources · checked on 9 October 2026. Availability and limits are listed below.

  • Plan work

    Google describes goal-based task planning.

    Source 2
  • Skills and tools

    Reusable skills connect workflow steps with tools.

    Source 1
  • Persistent execution

    Announced cloud execution keeps longer tasks running.

    Source 1
  • Permissions and logs

    Google describes agent identities, access policies and audit logs.

    Source 1
  • Generate code and content

    The announcement covers coding and content creation.

    Source 2
  • Coordinate temporary subagents

    Splits multi-step work across specialist subagents with their own identities.

    Source 1
  • Work as a persistent coworker

    Describes team agents with a dedicated role, email address and persistent storage.

    Source 1
  • Respond to schedules and events

    Start tasks on a schedule or in response to events; cloud context persists across devices.

    Source 1
  • Work in Workspace and other channels

    Announced interfaces include Workspace, Microsoft 365, Slack, CLI, apps and an API.

    Source 1
02

Requirements and limits

  • Not yet confirmed

    General access remains unconfirmed

    The release describes features and architecture but provides no complete rollout and pricing plan for all interfaces.

    Source 1
  • Requirement

    Dedicated identities for coworker agents

    Coworker agents have their own email addresses and storage. Their work context comes from information provided by people or teams.

    Source 1
  • Requirement

    Enterprise policies for tool access

    Identities, authorization, sandboxes and network gateways are the documented controls for access to enterprise systems.

    Source 1

Accounts & permissions

Accounts and permissions for this profile have not yet been researched.

Operation & oversight

Operational details for this profile have not yet been researched.

03

Practical tasks

Editorial examples based on documented features. No product tests performed.

01

Prepare a status report

Your input
Approved project documents and a reporting brief.
Possible result
Editorial example: a draft report.
What to check
Review sources, claims and recipients.
02

Delegate analysis

Your input
A bounded question and accessible data.
Possible result
Editorial example: an analyzed question.
What to check
Review data permissions, calculations and results.
EDITORIAL EXAMPLE

Editorial use case after access is enabled: review project documents and prepare a status report.

Your possible prompt

Compare three providers for [task]. Use their official documentation. Show differences, sources, and open questions in a table.

Getting-started idea based on documented capabilities. Not an executed product test. Adapt the task, data, and permissions.

04

Setup and pricing

Setup

Confirm rollout and enterprise access; configure data sources and permissions

Prices and included usage

Pricing checked

The universal Gemini agent was announced on 8 October. The announcement does not state a dedicated public plan.

  • Gemini Code Assist and Agent Platform prices cover other products and are not listed here as Gemini agent prices.

Pricing details

05

Sources and evidence

Official Google announcements dated October 8, 2026. Access and pricing for a specific deployment require confirmation.

Editorial team: Michael Meierhoff · sources reviewed on . Provider information is not independent evidence of performance.

Read profile as Markdown

  • cloud.google.com /blog/products/ai-machine-learning/welcome-to-gemini-at-work-2026
  • blog.google /innovation-and-ai/infrastructure-and-cloud/google-cloud/gemini-at-work/
How we assess the sources