Build secure, Google-quality enterprise-grade generative AI applications for search and recommendations. Agent Search lets developers, even those with limited machine learning skills, to tap into the power of Google's foundation models and search and recommendation expertise.

Try it!

Try this working demo of Agent Search, indexed on this documentation site. It retrieves search results and provides AI-generated answers based on the results. You can search with keywords or natural language queries.

To get an AI-generated answer, ask a complete question like What is Agent Search for media? or What are the features of custom search?

Key capabilities and configurations

Agent Search offers the following key capabilities and configurations:

  • Search

    Out-of-the-box natural language understanding and semantic search.

  • Recommendations

    ML-based content and metadata understanding so that your users quickly find content that is similar to the content that they're looking for.

  • Generative AI

    Generative AI-powered answers and summarization with conversational search.

  • Console or API

    Use the AI Application page in Google Cloud console or the Discovery Engine API to set up a search widget for your users.

  • Widget

    Integrate the search into your website using the widget.

  • Self-learning models

    Get self-learning ranking models and advanced analytics. This requires user events that based on user's clickstream.

  • Optimization for media

    Create recommendation and search apps optimized for media content.

  • Natural language querying of healthcare data

    Search FHIR resources without prior knowledge of any query language.

  • Context-aware healthcare searches

    Find search results with semantic relevance that a structured FHIR search might miss.

Get started

Agent Search offers the search and recommendations in the following business areas:
Made for vertical-agnostic use cases, use custom search to search within your public websites or to search within data stores containing your proprietary data.
Designed for media content such as movies, videos, and music, you can use media search to help your audience efficiently find the media content that they want to view or listen to.
A specialized search for healthcare records in FHIR data stores, allowing you to query both clinical data and referenced unstructured files like images and PDFs.
Powered by the user events that you gather, use custom recommendations to deliver the best results for your users.
Receive tailored recommendations for media content such as videos, news, and music, helping audiences discover personalized content with results customized by optimization objectives.

Product name terminology

Agent Search is a child product of AI Applications. Because this documentation set only discusses Agent Search. The name AI Applications is used on the Google Cloud Console UI and in the documentation to discuss the UI page.

Behind the scenes, Agent Search uses the Discovery Engine API. When you run curl commands or programmatically use the API, you see references to the discoveryengine.googleapis.com endpoint. The Gemini Enterprise product also uses the Discovery Engine API.

Go to AI Applications console