Vector database for RAG: what fits on your server

Your server RAM

Your server disk

Corpus text 2 GB

Embedding model (sets dimension)

Requirements

Options

Quantization (shrinks vectors)

Store chunk text (returned in RAG)

Databases that fit your server, most headroom first. Grey rows are ruled out: too large for your server, or missing a requirement you selected. Hover a row for exact RAM, cost, and notes.

DatabaseOn your serverMax it holds