Applied AI·Embeddings and retrieval
the embedding comes back as a fixed-length array of floats, and that length decides what it costs you to store and compare.
Vector / dimension
Also calleddimension
Draft summary, pending review
An embedding is a fixed-length array of floats; the length is its dimensionality, commonly 256 to 3072. Higher dimensions capture more nuance but cost more to store and compare. Some models let you truncate dimensions with modest quality loss.