mysql-vector

MySQL-Vector integrates native vector data storage (VECTOR type) and similarity search functions into the MySQL 9.3+ relational database engine.

This technology is the native `VECTOR` data type and associated functions introduced in MySQL 9.3. It allows developers to store high-dimensional numerical embeddings directly in a relational table, eliminating the need for a separate vector database. The `VECTOR(N)` type stores up to 16,383 single-precision (4-byte) floating-point values. Key functions like `DISTANCE()` enable efficient similarity search (e.g., COSINE, DOT, EUCLIDEAN metrics), directly supporting GenAI applications like Retrieval-Augmented Generation (RAG) and semantic search within the familiar MySQL ecosystem.

https://dev.mysql.com/doc/refman/9.3/en/vector.html