Gemini Enterprise AI FAQs

Below is a list of questions and answers for the Gemini Enterprise AI.

Is this service free for the Stanford community?: There are two tiers. The Standard tier is free to students, postdocs, faculty, and staff. The Pro tier has a monthly charge. Visit the Gemini Enterprise AI Rates webpage to learn more.
What connectors, plugins, and addons are available?: View the Approved Connectors for AI Services webpage to learn more.
Am I able to access Google Agent Space using Gemini Enterprise?: Yes. Google Agentspace is now part of Gemini Enterprise, with the related agent creation and orchestration technology now powering Gemini Enterprise.
Do I need to be a developer to use Gemini Enterprise?: No. Gemini Enterprise offers low code and no code options for building and deploying pre-made or custom agents.
How can I create an agent in Gemini Enterprise?: Gemini Enterprise provides centralized oversight and management for agents used by the university campus. You can learn more about this on the Google page titled, Use Gemini Enterprise > Create an agent.
Can I share an agent I created in Gemini Enterprise?: No. At this time you cannot share personal agents running in Gemini Enterprise. You can share the chat record, but not the agent itself.
Can I use Gemini Enterprise to generate video, audio, and images?: Yes. Gemini Enterprise includes video, audio, and image generation capabilities too. Please note that there are limits to how many of each type of media can be generated per day. Review the FAQ items about limits in Gemini Enterprise and NotebookLM Enterprise for more details.
How is Gemini Enterprise different from Google Gemini Chat or other AI services?: Gemini Enterprise focuses on AI agents and agentic capabilities. Where Google Gemini Chat allows you to chat with an AI model, Gemini Enterprise allows you to automate workflows, build tools, browser websites, take actions online, and conduct deep research.
Are there any limits to Gemini Enterprise or NotebookLM Enterprise?: Yes. Gemini Enterprise and NotebookLM each have limits on their usage. These limits can restrict how many queries, images, videos, deep research tasks, etc. you can perform each day and reset back to zero each day.

In addition to individual daily limits, Gemini Enterprise has pooled organizational limits too. For example, if an individual license has a limit of 10 queries per user per day, and you have 10 user licenses, the entire environment would have a total pooled quota of 100 queries per day (10 licenses x 10 queries). This does not apply to NotebookLM Enterprise.

As these limits are subject to change at any time by Google, we will link to Google's documentation about these limits below:
Learn more about Gemini Enterprise EDU limits.
Learn more about NotebookLM Enterprise limits.
Can I purchase NotebookLM Enterprise or NotebookLM Plus separately from Gemini Enterprise?: No. Google NotebookLM Plus and NotebookLM Enterprise cannot be purchased for use with Stanford accounts due to vendor licensing limitations. UIT will continue to work with Google to find solutions to this in the future.
Can I pay for the use of Vertex AI tokens to run ADK agents hosted on Vertex AI Agent Engine?: No. Due to limits in Google's billing model, UIT is unable to enable ADK agents by request. UIT will evaluate which ADK agents can be made available and will centrally fund their usage.
How does Gemini Enterprise AI handle sensitive data?: This is a core strength of the enterprise platform. Gemini Enterprise is part of Stanford's Google Cloud platform. All sensitive data stays within the Stanford environment. The tool has not yet been approved by ISO.
Is Gemini Enterprise FERPA compliant?: Yes. Based on current risk classifications, Stanford's Gemini Enterprise AI is approved for use with moderate risk data, like FERPA data.
Is Stanford data used to train Google AI models?: No. Data entered into the Gemini Enterprise and NotebookLM is not used to train models. The university has agreements in place which state that Google is prohibited from performing any custom training or fine tuning of models with Stanford data.

Billing

How will billing work in the new Gemini Enterprise AI Workspace plan?

There are two tiers:

Standard: Free to students, postdocs, faculty, and staff.
Pro: Has a monthly charge that requires a PTA.

What payment methods are required for the Pro tier?

A valid PTA is required to purchase the Pro tier. STAP funds cannot be used for this upgrade.

You can learn more about PTAs on Stanford's Admin Guide 3.1.3 Expenditure Accounts (PTAs) or through Stanford's Fingate page about PTA Manager system. The UIT AI Team will not be able to help you find or create a PTA.

Do I always need to provide PTA billing information for the Pro tier?

Yes. A valid PTA is required to purchase the Pro tier.

What does "Project Task Award" refer to on the Add Gemini Enterprise Service form?

A PTA is a method of labeling and tracking expenditure accounts in the university's financial systems. It consists of three segments: Project (P), Task (T), and Award (A), which help organize and allocate funds for various academic and research initiatives. You enter the corresponding sections into the fields in the request form.

(In the above example, let's use the fake PTA 5555555-000-APIGW. The first section of the PTA is the project, the second is the task, and the last is the award.)