Documentation | Vector Stores

Vector Stores

VectorStoresResource

Methods

client.vector_stores.list() -> SyncCursorPage[

VectorStore

]

get/v5/vector-stores

List all vector stores for the current account with pagination.

Returns a paginated list of vector stores owned by the authenticated account.

Parameters

ending_before: Optional[str]

limit: Optional[int]

(maximum: 10000, minimum: 1, default: 100)

sort_by: Optional[str]

sort_order: Optional[Literal["asc", "desc"]]

(default: "asc")

"asc"

"desc"

starting_after: Optional[str]

Returns

VectorStore

created_at:

datetime

(format: date-time)

Timestamp of creation

embedding_dimensions: int

Dimension size of the embeddings

embedding_model: Literal["sentence-transformers/all-MiniLM-L12-v2", "sentence-transformers/all-mpnet-base-v2", "sentence-transformers/multi-qa-distilbert-cos-v1", "sentence-transformers/paraphrase-multilingual-mpnet-base-v2", "openai/text-embedding-ada-002", "openai/text-embedding-3-small", "openai/text-embedding-3-large", "embed-english-v3.0", "embed-english-light-v3.0", "embed-multilingual-v3.0", "gemini/text-embedding-005", "gemini/text-multilingual-embedding-002", "gemini/gemini-embedding-001"]

The embedding model

name: str

The name of the vector store

updated_at:

datetime

(format: date-time)

Timestamp of last update

indexed_metadata_fields: Optional[Dict[str, Literal["string", "number", "boolean", "object", "list"]]]

Dictionary mapping metadata field names to their types

Request example

import os
from scale_gp_beta import SGPClient

client = SGPClient(
    api_key=os.environ.get("SGP_API_KEY"),  # This is the default and can be omitted
)
page = client.vector_stores.list()
page = page.items[0]
print(page.created_at)

200Example

{
  "has_more": true,
  "items": [
    {
      "created_at": "2019-12-27T18:11:19.117Z",
      "embedding_dimensions": 0,
      "embedding_model": "sentence-transformers/all-MiniLM-L12-v2",
      "name": "name",
      "updated_at": "2019-12-27T18:11:19.117Z",
      "indexed_metadata_fields": {
        "foo": "string"
      }
    }
  ],
  "total": 0,
  "limit": 0,
  "object": "list"
}

client.vector_stores.create() ->

VectorStore

post/v5/vector-stores/create

Create a new vector store.

The vector store will be scoped to the authenticated account. The name must be unique within the account.

Args: request: Vector store creation parameters including name, dimensions, and optional model vector_store_use_case: Injected vector store use case

Returns: The created vector store details

client.vector_stores.retrieve() ->

VectorStore

get/v5/vector-stores/{vector_store_name}

Get a vector store by name.

Args: vector_store_name: The unique name of the vector store within the account vector_store_use_case: Injected vector store use case

Returns: The vector store details

client.vector_stores.configure(, ) ->

VectorStore

post/v5/vector-stores/{vector_store_name}/configure

Configure the settings of a vector store.

Currently only supports updating indexed_metadata_fields. The name, embedding_dimensions, and embedding_model are immutable after creation.

Args: request: Configuration update with indexed_metadata_fields vector_store_name: The unique name of the vector store within the account vector_store_use_case: Injected vector store use case

Returns: The updated vector store details

client.vector_stores.drop() ->

VectorStoreDropResponse

post/v5/vector-stores/{vector_store_name}/drop

Delete (drop) a vector store by name.

This is a hard delete operation that permanently removes the vector store.

Args: vector_store_name: The unique name of the vector store within the account vector_store_use_case: Injected vector store use case

Returns: Deletion confirmation with the vector store ID

client.vector_stores.upsert(, ) ->

VectorStoreUpsertResponse

post/v5/vector-stores/{vector_store_name}/upsert

Insert or update documents in a vector store.

Existing documents (by ID) will be updated, new documents will be inserted.

Args: request: Array of documents to upsert vector_store_name: The unique name of the vector store vector_store_use_case: Injected vector store use case

Returns: Batch operation results with success/failure counts

client.vector_stores.delete(, ) ->

VectorStoreDeleteResponse

post/v5/vector-stores/{vector_store_name}/delete

Delete documents from a vector store by IDs or filter.

Args: request: Either IDs or filter criteria for deletion vector_store_name: The unique name of the vector store vector_store_use_case: Injected vector store use case

Returns: Number of documents deleted

client.vector_stores.vectors(, ) ->

VectorStoreVectorsResponse

get/v5/vector-stores/{vector_store_name}/vectors

List documents in a vector store with pagination.

Args: vector_store_name: The unique name of the vector store limit: Maximum number of documents per page (max 100 with vectors, 1000 without) cursor: Pagination cursor from previous response filter_param: Optional metadata filter (JSON string) include_vectors: Whether to include embedding vectors vector_store_use_case: Injected vector store use case

Returns: Paginated list of documents

client.vector_stores.count(, ) ->

VectorStoreCountResponse

post/v5/vector-stores/{vector_store_name}/count

Count documents in a vector store, optionally filtered by metadata.

Args: vector_store_name: The unique name of the vector store request: Optional filter criteria (empty body counts all documents) vector_store_use_case: Injected vector store use case

Returns: The count of documents matching the criteria

client.vector_stores.query(, ) ->

VectorStoreQueryResponse

post/v5/vector-stores/{vector_store_name}/query

Query documents by similarity search.

Supports semantic (vector), lexical (text), or hybrid search modes.

Args: request: Query parameters including text, filters, and reranking options vector_store_name: The unique name of the vector store vector_store_use_case: Injected vector store use case

Returns: Matching documents with similarity scores and query metadata

Domain types

class VectorStore: ...

Response model for vector store operations.