Skip to main contentDocumentation index
Fetch the complete documentation index at: /docs/llms.txt
Search it with GET /docs/search?q=<query>.
Use these to discover all available pages before exploring further.
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.
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.
The initial number of vectors to retrieve in the query result. The response will be sorted based on the distance metric score.
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.
Whether to include the vector values in the response.
Whether to include the data field in the response.
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
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.
Fusion algorithm to use while fusing scores from dense and sparse components of a hybrid index. If not provided, defaults to RRF.
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.
Maximum idle time for the resumable query in seconds.
The ID of the resulting vector.
The score of the vector data, calculated based on the distance metric of your index.
The resulting vector (if includeVectors is set to true)
The resulting sparseVector (if includeVectors is set to true)
The metadata of the vector (if includeMetadata is set to true)
The data of the vector (if includeData is set to true)
Was this page helpful?
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 resultsconsole.log(nextBatch);await stop(); // Stop the resumable queryconst { result, fetchNext, stop } = await index.resumableQuery({ maxIdle: 3600, topK: 50, data: "lord of the rings" includeMetadata: true, includeData: true,});console.log(result);/*[ { id: '6345', score: 1.00000012, data: "hobbit", metadata: { sentence: "Upstash is great." } }, // ... more results]*/const nextBatch = await fetchNext(5); // Fetch next 5 resultsconsole.log(nextBatch);await stop(); // Stop the resumable querytype Metadata = { title: string, genre: 'sci-fi' | 'fantasy' | 'horror' | 'action'}const { result, fetchNext, stop } = await index.resumableQuery<Metadata>({ vector: [ ... // query embedding ], includeMetadata: true, topK: 1, filter: "genre = 'fantasy' and title = 'Lord of the Rings'", maxIdle: 3600,})if (result[0].metadata) { // Since we passed the Metadata type parameter above, // we can interact with metadata fields without having to // do any typecasting. const { title, genre } = result[0].metadata; console.log(`The best match in fantasy was ${title}`)}await stop();