LaunchedEditorial Listing

Gemini for Business

Google · Gemini for Business: Google Cloud's Universal Work Agent, Sold as Gemini Enterprise

Open Gemini for Business

Gemini for Business is Google Cloud's enterprise work agent, sold as the Gemini Enterprise editions. Employees hand it questions, documents, data analysis and code from one prompt box. It works across Google Workspace, Microsoft 365, Slack and business systems such as Salesforce, Jira and BigQuery, under each user's existing permissions, with audit trails and per-project budget caps. It suits organizations that want one governed agent for staff instead of a separate assistant in every tool.

PricingPaid
Setupmedium
Runs onWeb
Open sourceNo
DocsYes
CategoryProductivity
EnterpriseAutonomous AgentCloud AgentsBackground AgentsSubagentsPersistent MemoryAgent SkillsMulti-ModelMCPGovernanceCost ControlGoogle WorkspaceMicrosoft 365Enterprise Search

Best for

Mid-size and large organizations, especially Google Cloud or Google Workspace customers, that want to give employees one governed agent for knowledge work, data analysis and light coding across Workspace, Microsoft 365, Slack and systems like Salesforce, Jira and BigQuery, with per-agent identity, audit trails and per-project spend caps

Not ideal for

Individuals wanting a consumer assistant, teams that need the newest agent capabilities on a firm schedule (several are early access), organizations that require self-hosted or air-gapped deployment, and developers who need a model API or agent framework rather than an employee-facing agent

Who it's for

Knowledge workers across sales, marketing, finance, legal, engineering and operations, rolled out by IT and security teams on Google Cloud

Capabilities

  • One prompt box for questions, knowledge work, image and media creation, data analysis and writing and running code, with objectives you delegate rather than step-by-step prompts
  • Scheduled tasks and event-triggered work, alongside on-demand chat
  • Long-running jobs execute on their own secure machine in Google Cloud and continue after the user logs off (multi-step background tasks are in early access)
  • Four kinds of memory carried across devices and channels: session, semantic (business knowledge), procedural (how work gets done, including skills it writes) and episodic (past actions)
  • Temporary sub-agents with their own identities for multi-step jobs, plus coworker agents with a persistent team role and their own @agents email, Workspace account, calendar and Drive, as described by Google
  • Skills and tools registries where teams publish reusable skills and tools company-wide, plus personal skills; Standard and Plus editions can create and upload custom skills
  • More than 100 integrations listed by Google, including Google Workspace, Microsoft 365 (Outlook, Teams, OneDrive, SharePoint), Slack, Confluence, Notion, Jira, Linear, Asana, GitHub, GitLab, Salesforce, HubSpot, ServiceNow, Docusign and BigQuery, plus custom MCP servers
  • Indexes selected sources while keeping data where it lives, with access following each user's existing permissions
  • Model routing across Gemini models and Anthropic's Claude models under one set of permissions and one bill (third-party model choice is in early access)
  • Per-agent identity with least-privilege permissions, an audit trail attributed to the agent, Agent Sandbox execution and Agent Gateway policy enforcement on agent traffic
  • Per-project hard spend caps set in the Cloud Billing Console: the agent pauses when a cap is reached and costs can be charged back by project
  • Prebuilt Google agents such as Deep Research and Gemini Notebook Enterprise, a no-code agent builder (preview), an Agent Marketplace, and publishing of agents built on Gemini Enterprise Agent Platform or ADK, or connected over A2A (Standard and Plus)
  • Industry specializations for Financial Services and Legal in preview, with Government, Healthcare and Retail announced as coming soon

Limitations

  • Several headline capabilities are early access, not available to all customers: delegating multi-step background tasks, Gemini on mobile and desktop, and third-party model choice
  • Google has not announced separate pricing for the new Gemini agent or said whether every capability is included in existing Gemini Enterprise subscriptions
  • Only the Business edition has a published starting price ($21 per seat per month); Standard and Plus seat prices are not listed on Google's pages, so buyers go through the Cloud console, a free trial or sales
  • Seat-based editions are bounded by pooled daily quotas (for example Assistant queries and Deep Research runs), with overages billed at pay-as-you-go rates only if enabled
  • Self-serve console purchases are capped at 25 seats for online billing accounts and 1,000 for invoiced accounts; larger orders go through sales
  • Runs only as a Google Cloud service: no self-hosted or on-premises deployment is documented
  • Building custom code-based agents requires the separately purchased Gemini Enterprise Agent Platform, which uses pay-as-you-go billing

Use cases

  • Delegating a quarter-end financial review that pulls data from BigQuery and Sheets and returns a model and slide deck
  • Preparing for a sales meeting by gathering CRM history, email threads and competitor research, then drafting follow-up notes
  • Setting up a coworker agent for a team, such as an events coordinator that drafts launch-readiness documents when @mentioned in Chat
  • Asking plain-language questions of governed warehouse data and saving the resulting query as a reusable report
  • Running a long research or analysis job overnight on a cloud machine and reviewing the finished output the next morning (multi-step background tasks are in early access)
  • Rolling out one agent to a workforce while capping AI spend per project and charging costs back to departments

Our take

Google's pitch is consolidation. Instead of a separate assistant in every SaaS tool, you deploy one agent that carries memory, skills and permissions across Workspace, Microsoft 365, Slack and systems of record. It is the governance layer, not the chat, that makes this credible: agent identities, attributable audit logs, sandboxed code execution, gateway policies and hard per-project spend caps. Those answer the questions that usually stall enterprise agent rollouts. The catch is timing and transparency. The most distinctive pieces, including delegated background tasks, Claude routing and the desktop and mobile apps, are early access, and Google has not said what the new agent costs on top of existing editions. Evaluate it now if you already run on Google Cloud or Workspace and want a single governed agent. If you are standardized on Microsoft 365 or need to self-host, compare it with suite-native or privately deployable options first.

Who should use it

IT, security and AI program leaders at organizations on Google Cloud or Google Workspace, or running mixed Google and Microsoft estates, who want employees to delegate research, documents, data analysis and code to one agent, with agent identities, audit trails and project-level budget controls in place.

Who should skip it

Individuals and very small teams looking for a consumer assistant, organizations that must self-host or run air-gapped, buyers who need firm published pricing before evaluating, and developers who want a model API or agent framework rather than an employee-facing agent.

Strengths

  • One agent for questions, documents, data analysis, media and code instead of separate assistants per tool
  • Long-running jobs on cloud machines plus persistent memory across devices and channels
  • Broad connector list spanning Google Workspace and Microsoft 365, Slack, Atlassian, Salesforce, ServiceNow and BigQuery, plus MCP
  • Per-agent identity, audit trails attributed to agents, sandboxed execution and gateway policies
  • Per-project hard spend caps with cost chargeback by project
  • Routes jobs across Gemini and Claude models under one bill

Weaknesses

  • Background tasks, mobile and desktop apps, and third-party model choice are early access
  • No published price for the new agent; Standard and Plus seat prices are not listed
  • Pooled per-edition quotas with overage billing to manage
  • Google Cloud-hosted only; no self-hosted option
  • Custom code agents need the separately billed Agent Platform

Gemini for Business pricing

Gemini Enterprise Business

From $21 per seat

Billed monthly

  • For individuals, small businesses and teams, no IT setup needed
  • 25 GiB pooled storage and data indexing per user
  • Select business connectors, Gemini Notebook, Deep Research and no-code Workflow Builder
  • 30-day free trial

Gemini Enterprise Standard

Per seat (price not published)

Billed monthly

  • Monthly or annual subscription; 30-day free trial
  • 30 GiB pooled storage per user; 160 Assistant queries and 3 Deep Research runs per user per day, pooled
  • Full connector ecosystem, Agent Marketplace, custom skills and bring-your-own agents
  • $10 monthly credit per user for Antigravity and Android Studio (invoiced billing accounts)

Gemini Enterprise Plus

Per seat (price not published)

Billed monthly

  • Monthly or annual subscription
  • 75 GiB pooled storage per user; 200 Assistant queries and 10 Deep Research runs per user per day, pooled
  • 30-day no-cost trial without sales
  • $15 monthly credit per user for Antigravity and Android Studio (invoiced billing accounts)

Gemini Enterprise Pay-as-you-go

Usage-based

  • No feature quota limits; pay for what you use
  • Requires an invoiced Cloud Billing account
  • Storage and data indexing billed at $5 per GiB per month

Gemini Enterprise Frontline

Contact sales

  • Add-on for frontline workers
  • Requires 150 or more Standard or Plus users
  • 2 GiB pooled storage per user

Note: Gemini for Business is sold as the Gemini Enterprise editions. Google's Business edition page lists that edition from $21 per seat per month; Standard and Plus seat prices are not published on Google's product, FAQ or editions pages. Google has not announced separate pricing for the Gemini agent introduced on October 8, 2026, or said whether every capability is included in existing subscriptions. Feature quotas are pooled per edition within a project and location, and AI developer tool credits apply only to invoiced Cloud Billing accounts. When pooled quotas run out, overages can be enabled at pay-as-you-go rates. Admins can set hard per-project AI spend caps in the Cloud Billing Console. Gemini Enterprise Agent Platform, for building custom agents, is billed separately on a pay-as-you-go basis.

Technical specs

Available models

Gemini models (including Gemini 3.8 Flash, enabled by default, and Gemini 2.5 Pro)Anthropic Claude models (third-party model choice in early access)

Where Gemini for Business excels

Scheduling a cross-team meeting without naming attendees

Google's example has Gemini identify the usual regional event leads from a chat space and a previous thread, check their calendars and start an email thread to agree a time, including with external participants.

A shared coworker agent for a marketing team

A coworker agent gets its own Workspace account and can be @mentioned in Chat or tagged in a Doc comment. It drafts documents, suggests edits under its own name and sees only what the team shares with it.

Self-serve operational reporting for business users

Gemini uses BigQuery and Knowledge Catalog to build and save a query from a plain-language request, and teams can re-run saved reports on demand without incurring token costs.

Gemini for Business vs. competitors

Gemini for Business vs. Microsoft Copilot

Microsoft 365 Copilot is built into Word, Excel, PowerPoint, Outlook and Teams and grounded in Microsoft 365 work data. Gemini for Business is a Google Cloud agent that works inside Google Workspace and also connects to Microsoft 365, Slack, Salesforce, ServiceNow and BigQuery, routes across Gemini and Claude models, and adds per-project spend caps. Its new agent pricing is unannounced, and several of its features are early access.

Gemini for Business vs. ChatGPT

ChatGPT Business and Enterprise give employees OpenAI's models with connectors, agentic features and admin controls as a hosted workspace. Gemini for Business is sold through Google Cloud, ties agent identity, audit logs and spend caps into Google Cloud's IAM and billing, and can route work to Claude as well as Gemini models.

Gemini for Business vs. Cohere North

Cohere North is a model-agnostic agent workspace that can be deployed in your VPC, on-premises or air-gapped, with custom pricing. Gemini for Business runs only as a Google Cloud service but brings Google's models, Workspace integration, coworker agents with their own Workspace accounts, and Cloud Billing spend caps.

Gemini for Business vs. Claude Cowork

Claude Cowork is Anthropic's task-execution mode inside the Claude app on paid plans, working across your files and connected tools with Claude models. Gemini for Business is an organization-wide Google Cloud agent with sub-agents and coworker agents, company skills and tools registries, per-agent identity and audit, and routing that includes Claude models alongside Gemini.

Frequently asked questions

What is Gemini for Business?

It is Google Cloud's enterprise work agent, presented on Google's product page as "Gemini for Business: The Universal Work Agent". Employees use one prompt box to ask questions, delegate knowledge work, analyze data, create media and write and run code across their email, documents, code, data and business apps, under their existing permissions.

Is Gemini for Business the same as Gemini Enterprise or Google Agentspace?

It is the same product line. The product page at Google's Gemini Enterprise URL now uses the Gemini for Business name, while editions, documentation and billing still use Gemini Enterprise (Business, Standard, Plus, Pay-as-you-go and Frontline editions). Google Agentspace became a feature of Gemini Enterprise on October 9, 2025, and is no longer sold as a standalone service.

How is it different from the Gemini app and Gemini Spark?

The Gemini app is Google's general AI assistant, included with Google Workspace business plans and offered to consumers. Gemini Spark is a personal agent for Google AI subscribers. Gemini for Business is sold through Google Cloud and adds connectors to non-Google systems, custom and third-party agents, per-agent identity, audit trails and project-level spend controls.

How does it relate to Gemini Enterprise Agent Platform?

Gemini Enterprise Agent Platform, the evolution of Vertex AI, now shown as Gemini platform, is the developer platform for building, scaling and governing custom agents. Google sells it separately with pay-as-you-go billing. Agents built there or with the Agent Development Kit can be published into Gemini for Business on Standard and Plus editions.

How much does Gemini for Business cost?

It is sold as per-seat Gemini Enterprise editions plus a usage-based Pay-as-you-go edition. Google lists the Business edition, for small businesses and teams, from $21 per seat per month with a 30-day free trial. Standard and Plus seat prices are not published; they are bought as monthly or annual subscriptions in the Cloud console or through sales, and Gemini Enterprise Plus can be tried at no cost for 30 days without sales. Google has not announced separate pricing for the new Gemini agent.

Which AI models does it use?

Google says it routes each job to a suitable model across the Gemini family and Anthropic's Claude models, with more models coming. In the Gemini Enterprise app, Gemini 3.8 Flash is generally available and enabled by default, and admins can also turn Gemini 2.5 Pro on or off. Third-party model choice is currently in early access.

Does Google train models on my company's data?

Google says customer data, including prompts, outputs and model tuning, is never used to train Google models or models for other customers, is not sold to third parties and is not used for advertising.

Gemini for Business vs Microsoft 365 Copilot: what is the difference?

Microsoft 365 Copilot works inside Word, Excel, PowerPoint, Outlook and Teams, grounded in Microsoft 365 data. Gemini for Business is a Google Cloud agent that also works inside Google Workspace but connects to Microsoft 365 apps, Slack, Salesforce, ServiceNow, Jira and BigQuery. It routes across Gemini and Claude models and adds per-project spend caps in Cloud Billing.

Integrations & fit

Google WorkspaceGmailGoogle DriveGoogle CalendarGoogle ChatMicrosoft 365Microsoft OutlookMicrosoft TeamsMicrosoft SharePointMicrosoft OneDriveSlackConfluenceNotionJiraLinearAsanaGitHubGitLabSalesforceHubSpotServiceNowDocusignZendeskBigQueryMCP serversAgent2Agent (A2A)Gemini Enterprise Agent PlatformAgent Development Kit (ADK)
Good fit forStartup / small team, Enterprise
Pricing modelPaid· Paid subscription required
See pricing on Gemini for Business →

Alternatives to consider

About Gemini for Business

Gemini for Business is the name Google Cloud's product page now uses for the offering that is still sold, documented and billed as Gemini Enterprise. Gemini Enterprise launched in October 2025, and on October 9, 2025 Google Agentspace became a feature of it. At Gemini at Work on October 8, 2026, Google introduced the Gemini agent inside it: a single agent that answers questions, handles knowledge work, creates images and media, and writes and runs code. You give it an objective rather than step-by-step instructions, and you can also schedule tasks or have it respond to events. Google describes it as cloud-resident: lengthy jobs run on their own secure machine in Google Cloud and keep going after you log off, and the agent keeps session, semantic, procedural and episodic memory across devices and channels. For multi-step jobs it can create temporary sub-agents, each with its own identity. Google also describes coworker agents that hold a persistent team role, with their own @agents email address, Workspace account, calendar and Drive. Reusable skills and tools can be published to company-wide registries, and custom tools connect through MCP. The product page lists more than 100 integrations, including Workspace, Microsoft 365, Slack, Confluence, Notion, Jira, GitHub, Salesforce, HubSpot, ServiceNow and BigQuery, and data stays where it lives under each user's permissions. Governance is the main selling point: every agent has its own identity and least-privilege permissions, every action goes into an audit trail attributed to the agent, code runs in an Agent Sandbox, and traffic passes through Agent Gateway policies. Admins can also set a hard spend cap per project in the Cloud Billing Console, and the agent pauses when it is hit. Routing picks between Gemini models and Anthropic's Claude models for each job. It is not the same product as the consumer Gemini app or Gemini Spark, and it is not the developer-facing Gemini Enterprise Agent Platform (the evolution of Vertex AI, now shown as "Gemini platform"). That platform is bought separately on pay-as-you-go billing, and agents built on it or with ADK can be published into Gemini for Business. The tradeoffs: several launch features are still early access, including multi-step background tasks, the mobile and desktop apps, and third-party model choice. Google has not announced separate pricing for the new agent. The Business edition is listed from $21 per seat per month, but Standard and Plus seat prices are not published on Google's pages.

Updates from Gemini for Business

LaunchGemini agent introduced at Gemini at Work 2026

Google Cloud introduced the Gemini agent, a single agent for questions, knowledge work, media and code with persistent memory, sub-agents and coworker agents, long-running jobs on cloud machines, skills and tools registries, Gemini and Claude model routing, and per-project spend caps. The product page now presents the offering as Gemini for Business. Background tasks, mobile and desktop apps, and third-party model choice are in early access; Financial Services and Legal specializations are in preview.

New FeatureFederated query mode for Data Cloud connectors (Preview)

Gemini Enterprise added a preview mode that queries AlloyDB, BigQuery, Cloud SQL and Spanner in place over MCP with each user's own credentials, with Knowledge Catalog integration for business and technical context.

New FeatureGranular access controls for apps and data stores

Administrators can grant Gemini Enterprise Admin, Viewer or User roles on individual apps and data stores instead of across a whole Google Cloud project.

Are you the founder? Claim this listing →