Databricks Vector Search Cost

Databricks Vector Search costs depend on the endpoint type, indexed data, embedding generation, query volume, and storage.

Creating or updating an index can consume compute as documents are embedded and synchronized. Query cost depends on the endpoint and request volume. A small test index may not represent the cost of processing millions of rows.

Measure index size, synchronization frequency, embedding requests, query volume, and endpoint runtime. Include the cost of the model that generates embeddings when estimating total cost. Databricks documents Vector Search.