What is a Vector Database?
Vector databases are designed to store, manage, and search high-dimensional data called vector embeddings. Unlike traditional databases that query for exact matches, vector databases excel at finding items based on semantic similarity or "closeness." This is powerful for applications like AI-powered search, recommendations, and image recognition. This interactive tool simplifies the vector space into 2D to help you visualize how Create, Read, Update, and Delete (CRUD) operations work conceptually.
Simulated 2D Vector Space
Database State
Vectors currently in the database.
Database is empty. Use the 'Create' panel to add vectors.
Create (or Upsert) a Vector
Add a new item to the database. An "embedding model" (simulated here) converts your text into a vector, which is then stored with its ID and metadata. Most vector DBs use an "upsert" operation: if the ID exists, it's updated; otherwise, a new entry is created.
Conceptual Code