Google Vertex Vector Search icon

Google Vertex Vector Search

Interact with Google Vertex AI Vector Search

Actions9

Index → Create

AI-generated

Summary

Create a new vector search index in a specified Google Vertex AI location with a given name, embedding dimensions, and distance measure type.

Inputs

  • Location (required) — The geographic location where the Vector Search index will be created.
  • Index Name (required) — The name to assign to the new vector search index.
  • Dimensions (required) — The number of dimensions of the embedding vectors for the index.
  • Distance Measure Type (required) — The distance metric used for nearest neighbor search, such as cosine distance or L1 distance.

Output shape

a single JSON object representing the created index resource returned from Google Vertex AI's IndexServiceClient

The output includes details of the created index and may be used in subsequent operations. Errors during creation throw exceptions unless 'Continue On Fail' is enabled.

Examples

Example 1: Create an index named 'product-search-index' in 'us-central1' with 768 embedding dimensions and cosine distance, then use the returned index resource in later workflow steps.

Set 'Location' to 'us-central1', 'Index Name' to 'product-search-index', 'Dimensions' to 768, and 'Distance Measure Type' to 'COSINE_DISTANCE'.

Discussion