Python: Create a vector index

Creates a new vector index in a vector bucket.

• This is an alpha feature and may change in the future.

Parameters

• bucket_name

  (Required)

  The name of the vector bucket.

• index_name

  (Required)

  The name of the index to create.

• dimension

  (Required)

  The dimension of the vectors in this index.

• distance_metric

  (Required)

  The distance metric to use. Options: `cosine`, `euclidean`.

• data_type

  (Required)

  The data type of the vectors.

• metadata

  (Optional)

  Optional metadata configuration for the index.

Examples

Create vector index

supabase.storage.vectors().from_("vectors-bucket").create_index(
    index_name="my-index",
    dimension=128,
    distance_metric="cosine",
    data_type="float32"
)