Resumable Query - Upstash Documentation

The resumableQuery method allows you to perform queries that can be resumed to fetch additional results. This is particularly useful for large result sets or when implementing pagination.

The dimension of the query vector must match the dimension of your index.

The score returned from query requests is a normalized value between 0 and 1, where 1 indicates the highest similarity and 0 the lowest regardless of the similarity function used.

Arguments

Payload

ResumableQueryPayload

required

vector | sparseVector | data

number[] | SparseVector | string

required

There are two ways to use the resumableQuery method. You can either create the vectors on your own and pass directly the vector or sparseVector field, depending on your index type. Or you can pass the data field and create the embeddings using Upstash Embedding.

The query data/vector that you want to search for in the index.

topK

number

required

The initial number of vectors to retrieve in the query result. The response will be sorted based on the distance metric score.

includeMetadata

boolean

Whether to include the metadata of the vectors in the response. Setting this true would be the best practice, since it will make it easier to identify the vectors.

includeVectors

boolean

Whether to include the vector values in the response.

includeData

boolean

Whether to include the data field in the response.

filter

string

The metadata filtering of the vector. This is used to query your data based on the filters and narrow down the query results.

If you want to learn more about filtering check: Metadata Filtering

weightingStrategy

WeightingStrategy

For sparse vectors, what kind of weighting strategy should be used while querying the matching non-zero dimension values of the query vector with the documents. If not provided, no weighting will be used.

fusionAlgorithm

FusionAlgorithm

Fusion algorithm to use while fusing scores from dense and sparse components of a hybrid index. If not provided, defaults to RRF.

queryMode

QueryMode

Query mode for hybrid indexes with Upstash-hosted embedding models. Specifies whether to run the query in only the dense index, only the sparse index, or in both. If not provided, defaults to HYBRID.

maxIdle

number

Maximum idle time for the resumable query in seconds.

Response

ResumableQueryResponse

Object

required

result

QueryResult[]

required

The initial query results.

fetchNext

(additionalK: number) => Promise<Vector[]>

required

A function to fetch the next batch of results.

stop

() => Promise<string>

required

A function to stop the resumable query and release resources.

QueryResult

Object

required

string | number

required

The ID of the resulting vector.

score

number

required

The score of the vector data, calculated based on the distance metric of your index.

vector

number[]

The resulting vector (if includeVectors is set to true)

sparseVector

SparseVector

The resulting sparseVector (if includeVectors is set to true)

metadata

Record<string, unknown>

The metadata of the vector (if includeMetadata is set to true)

data

string

The data of the vector (if includeData is set to true)

const { result, fetchNext, stop } = await index.resumableQuery({
  maxIdle: 3600,
  topK: 50,
  vector: [0, 1, 2, ..., 383], // 384-dimensional vector
  includeMetadata: true,
  includeVectors: true,
});

console.log(result);
/*
[
  {
    id: '6345',
    score: 1.00000012,
    vector: [0, 1, 2, ..., 383],
    metadata: {
      sentence: "Upstash is great."
    }
  },
  // ... more results
]
*/

const nextBatch = await fetchNext(5); // Fetch next 5 results
console.log(nextBatch);

await stop(); // Stop the resumable query

​Arguments

​Response

Arguments

Response