How to Restrict Gemini Grounding Search to Specific Domain(s)?

I’m currently experimenting with Gemini’s grounding search capabilities and have a specific use case in mind.

I’d like Gemini to perform retrieval only within a defined set of domains (eg: example1[dot]com, example2[dot]com, example3[dot].com etc.) , excluding any external sources or general web search results.

Ideally, Gemini should be able to access and retrieve information from all publicly available content within those domains, including internal pages and sublinks.

In short, I want Gemini to function as a domain-specific search engine powered by its grounding capabilities.

The The URL context tool is only valuable for a particular link and not for the entire domain. So this doesnt serve the purpose.

What’s the official way to configure Gemini’s grounding search to achieve this?
Any other workarounds for this??

Hi @vishalanand.ai
Welcome to the AI Forum!!!

You can connect your Gemini model to your website or document data using Vertex AI Search. Start by setting up a Vertex AI Search app with your domain URL. This is the best option for grounding Gemini across an entire domain.
For external websites, domain verification is required.