Topics / #image-search

Image-search projects on GitHub

Searching images by visual similarity using vector embeddings.

As of , the most-starred Image-search projects on GitHub are milvus (46.4K stars), qdrant (35.1K stars) and weaviate (16.9K stars). GitHub.Locker tracks 5 Image-search repositories with 114K stars combined. The fastest-growing on October 16 was qdrant, with +22 stars in a day.

5Projects
114KStars in total
+55Stars in a day

Hot right now

Most stars gained in a day · Oct 16

The all-time top

5 projects · by stars

  1. #1 · milvus-io/milvus

    milvus

    Milvus is a high-performance, cloud-native vector database built to store and search embedding vectors at scale for AI applications like semantic search, RAG, and recommendation systems.

    46.4K
  2. #2 · qdrant/qdrant

    qdrant

    Qdrant is a high-performance vector similarity search engine and vector database written in Rust, built to store, search, and manage 'points' (vectors plus a JSON payload) for AI applications like semantic search, recommendations, and RAG.

    35.1K
  3. #3 · weaviate/weaviate

    weaviate

    Weaviate is an open-source, cloud-native vector database: it stores objects and vectors together and lets you combine vector search with structured filtering for semantic search at scale.

    16.9K
  1. 4

    lancedb

    11.7Klancedb/lancedb · Rust

    LanceDB is an open-source vector database for multimodal AI: a central platform where you store, index and search petabytes of vectors and multimodal data, built for fast, production-ready retrieval.

  2. 5

    lap

    3.9Kjulyx10/lap · Vue

    Lap is an open-source, local-first photo manager for macOS, Windows, and Linux that lets you browse family albums, find old photos quickly, and organize large personal media libraries entirely offline.

Often tagged together