Class OPSQLiteVectorStore

SQLite-backed implementation of VectorStore using @op-engineering/op-sqlite.

Stores embeddings as an F32 blob and uses vector indexing helpers for efficient cosine-similarity search. Suitable for on-device retrieval in React Native apps.

Example

const store = await new OPSQLiteVectorStore({ name: 'vector-db', embeddings }).load();
await store.add({ document: 'hello world' });
const [top] = await store.query({ queryText: 'hello', nResults: 1 });
console.log(top.id, top.similarity);

Implements

VectorStore

Index

Constructors

constructor

Properties

Methods

add delete deleteVectorStore load query unload update

Constructors

constructor

new OPSQLiteVectorStore(
__namedParameters: { embeddings: Embeddings; name: string },
): OPSQLiteVectorStore
Creates a new SQLite-backed vector store.
Parameters
- __namedParameters: { embeddings: Embeddings; name: string }
Returns OPSQLiteVectorStore

Properties

db

db: DB

Methods

add

add(
    params: {
        document?: string;
        embedding?: number[];
        id?: string;
        metadata?: Record<string, any>;
    },
): Promise<string>
Inserts a document with an embedding. Generates an ID when not provided.
Parameters
- params: {
      document?: string;
      embedding?: number[];
      id?: string;
      metadata?: Record<string, any>;
  }
  Parameters for the operation.
  - Optionaldocument?: string
    Raw text content for the document.
  - Optionalembedding?: number[]
    Embedding for the document. If not provided, it will be generated based on the document.
  - Optionalid?: string
    ID for the document. If not provided, it will be auto-generated.
  - Optionalmetadata?: Record<string, any>
    Metadata for the document.
Returns Promise<string>
Promise that resolves to the ID of the newly added document.

Implementation of VectorStore.add

delete

delete(params: { predicate: (value: GetResult) => boolean }): Promise<void>
Deletes documents by predicate.
Parameters
- params: { predicate: (value: GetResult) => boolean }
  Parameters for deletion.
  - predicate: (value: GetResult) => boolean
    Predicate to match documents for deletion.
Returns Promise<void>
Promise that resolves once the documents are deleted.

Implementation of VectorStore.delete

deleteVectorStore

deleteVectorStore(): Promise<void>
Drops the vectors table entirely.

Returns Promise<void>

load

load(): Promise<OPSQLiteVectorStore>
Opens the database, loads embeddings, and ensures schema/indexes exist.

Returns Promise<OPSQLiteVectorStore>
Promise that resolves to the same instance.

Implementation of VectorStore.load

query

query(
    params: {
        nResults?: number;
        predicate?: (value: QueryResult) => boolean;
        queryEmbedding?: number[];
        queryText?: string;
    },
): Promise<QueryResult[]>
Executes a cosine-similarity query over stored vectors. Provide exactly one of queryText or queryEmbedding.
Parameters
- params: {
      nResults?: number;
      predicate?: (value: QueryResult) => boolean;
      queryEmbedding?: number[];
      queryText?: string;
  }
  Query parameters.
  - OptionalnResults?: number
    Number of top results to return.
  - Optionalpredicate?: (value: QueryResult) => boolean
    Function to filter results after retrieval.
  - OptionalqueryEmbedding?: number[]
    Precomputed query embedding.
  - OptionalqueryText?: string
    Raw query string to search for.
Returns Promise<QueryResult[]>
Promise that resolves to an array of QueryResult.

Implementation of VectorStore.query

unload

unload(): Promise<void>
Closes embeddings and the underlying database connection.

Returns Promise<void>
Promise that resolves when unloading is complete.

Implementation of VectorStore.unload

update

update(
    params: {
        document?: string;
        embedding?: number[];
        id: string;
        metadata?: Record<string, any>;
    },
): Promise<void>
Updates a document by ID.
Parameters
- params: {
      document?: string;
      embedding?: number[];
      id: string;
      metadata?: Record<string, any>;
  }
  Parameters for the update.
  - Optionaldocument?: string
    New content for the document.
  - Optionalembedding?: number[]
    New embeddings for the document. If not provided, it will be generated based on the document.
  - id: string
    ID of the document to update.
  - Optionalmetadata?: Record<string, any>
    New metadata for the document.
Returns Promise<void>
Promise that resolves when the update completes.

Implementation of VectorStore.update

Class OPSQLiteVectorStore

Example

Implements

Index

Constructors

Properties

Methods

Constructors

constructor

Parameters

Returns OPSQLiteVectorStore

Properties

db

Methods

add

Parameters

Optionaldocument?: string

Optionalembedding?: number[]

Optionalid?: string

Optionalmetadata?: Record<string, any>

Returns Promise<string>

delete

Parameters

predicate: (value: GetResult) => boolean

Returns Promise<void>

deleteVectorStore

Returns Promise<void>

load

Returns Promise<OPSQLiteVectorStore>

query

Parameters

OptionalnResults?: number

Optionalpredicate?: (value: QueryResult) => boolean

OptionalqueryEmbedding?: number[]

OptionalqueryText?: string

Returns Promise<QueryResult[]>

unload

Returns Promise<void>

update

Parameters

Optionaldocument?: string

Optionalembedding?: number[]

id: string

Optionalmetadata?: Record<string, any>

Returns Promise<void>

Settings

On This Page

`Optional`document?: string

`Optional`embedding?: number[]

`Optional`id?: string

`Optional`metadata?: Record<string, any>

`Optional`nResults?: number

`Optional`predicate?: (value: QueryResult) => boolean

`Optional`queryEmbedding?: number[]

`Optional`queryText?: string

`Optional`document?: string

`Optional`embedding?: number[]

`Optional`metadata?: Record<string, any>