Milvus Demo Hub: Explore AI-Powered Vector Search in Action

Curious about vector search? Try these hands-on demos to see how Milvus powers AI-driven image search, multimodal retrieval, RAG (Retrieval-Augmented Generation), and chemical structure search.

  • Multi-Phase Reranking with Function Chain

    Multi-Phase Reranking with Function Chain

    Build customized ranking logic in e-commerce with XGBoost Reranker and Function Chain in Milvus 3.0.

  • Data Curation in Autonomous Driving with StructArray

    Data Curation in Autonomous Driving with StructArray

    Store the nested object of a video clip and multiple key frames as a StructArray, together with labels at both levels, and search either level flexibly.

  • Visual PDF Retrieval with EmbeddingList

    Visual PDF Retrieval with EmbeddingList

    Natively implement multi-vector retrieval in Milvus with a ColPali-style model that ranks NASA handbook pages by MAX_SIM.

  • Multi-Modal Image Search

    Multi-Modal Image Search

    Visual semantic search with combined query image and text instructions.

  • Retrieval-Augmented Generation (RAG)

    Retrieval-Augmented Generation (RAG)

    Ask AI is a RAG chatbot for Milvus documentation and help articles. The vector database powering retrieval is Zilliz Cloud (fully-managed Milvus).

  • Chemical Structure Search

    Chemical Structure Search

    Blazing fast similarity search, substructure search, or superstructure search for a specified molecule.