Implementation:FlowiseAI Flowise QueryVectorStore

Attribute	Value
Sources	packages/ui/src/api/documentstore.js:L29
Domains	Document_Store_Ingestion
Last Updated	2026-02-12 14:00 GMT

Overview

QueryVectorStore is the API client function that sends a natural language query to the server for embedding and similarity search against a document store's configured vector store. It returns matching document chunks along with timing metrics for retrieval performance assessment.

Code Reference

Source Location

• File: packages/ui/src/api/documentstore.js, Line 29
• Repository: FlowiseAI/Flowise

Signature

const queryVectorStore = (body) => client.post(`/document-store/vectorstore/query`, body)

This function sends a POST request to the /api/v1/document-store/vectorstore/query endpoint via the configured axios client instance.

Import

import documentStoreApi from '@/api/documentstore'

The function is accessed as documentStoreApi.queryVectorStore(body) after importing the default export from the documentstore API module.

I/O Contract

Inputs

Parameter	Type	Required	Description
body	object	Yes	The query configuration
body.query	string	Yes	The natural language query text to search for
body.storeId	string	Yes	The document store identifier to search within
body.inputs	object	Yes	Vector store query configuration
body.inputs.topK	number	No	Maximum number of results to return (default varies by provider)
body.inputs.searchType	string	No	Search algorithm: "similarity" or "mmr" (Maximal Marginal Relevance)
body.inputs.filters	object	No	Metadata filters to constrain search results

Outputs

Field	Type	Description
data	object	The query result object
data.docs	Document[]	Array of matching document chunks, ordered by relevance
data.docs[].pageContent	string	The text content of the matching chunk
data.docs[].metadata	object	Metadata associated with the chunk (source, page, etc.)
data.timeTaken	number	Time in milliseconds for the query execution

The function returns a Promise<{data: {docs: Document[], timeTaken: number}}> from the axios response.

Usage Examples

Basic Similarity Search

import documentStoreApi from '@/api/documentstore'

const searchDocuments = async (storeId, queryText) => {
    const body = {
        storeId,
        query: queryText,
        inputs: {
            topK: 5,
            searchType: 'similarity'
        }
    }

    const response = await documentStoreApi.queryVectorStore(body)
    const { docs, timeTaken } = response.data

    console.log(`Found ${docs.length} results in ${timeTaken}ms`)
    docs.forEach((doc, i) => {
        console.log(`\n--- Result ${i + 1} ---`)
        console.log(`Content: ${doc.pageContent.substring(0, 200)}...`)
        console.log(`Source: ${doc.metadata.source || 'unknown'}`)
    })

    return response.data
}

MMR Search for Diverse Results

import documentStoreApi from '@/api/documentstore'

const diverseSearch = async (storeId, queryText) => {
    const body = {
        storeId,
        query: queryText,
        inputs: {
            topK: 8,
            searchType: 'mmr'  // Maximal Marginal Relevance for diverse results
        }
    }

    const response = await documentStoreApi.queryVectorStore(body)
    console.log(`MMR search: ${response.data.docs.length} diverse results in ${response.data.timeTaken}ms`)
    return response.data
}

Query with Metadata Filters

import documentStoreApi from '@/api/documentstore'

const filteredSearch = async (storeId, queryText, sourceFilter) => {
    const body = {
        storeId,
        query: queryText,
        inputs: {
            topK: 5,
            searchType: 'similarity',
            filters: {
                source: sourceFilter  // Only search chunks from a specific source
            }
        }
    }

    const response = await documentStoreApi.queryVectorStore(body)
    return response.data
}

Retrieval Quality Testing

import documentStoreApi from '@/api/documentstore'

const testRetrievalQuality = async (storeId, testQueries) => {
    const results = []

    for (const testQuery of testQueries) {
        const body = {
            storeId,
            query: testQuery.question,
            inputs: { topK: 5, searchType: 'similarity' }
        }

        const response = await documentStoreApi.queryVectorStore(body)
        const { docs, timeTaken } = response.data

        // Check if expected content appears in results
        const foundExpected = docs.some(doc =>
            doc.pageContent.includes(testQuery.expectedSnippet)
        )

        results.push({
            question: testQuery.question,
            resultCount: docs.length,
            timeTaken,
            foundExpected,
            topResult: docs[0]?.pageContent.substring(0, 100)
        })

        console.log(`"${testQuery.question}" => ${foundExpected ? 'PASS' : 'FAIL'} (${timeTaken}ms)`)
    }

    return results
}

Related Pages

• Principle:FlowiseAI_Flowise_Vector_Store_Query
• Implementation:FlowiseAI_Flowise_InsertIntoVectorStore -- Upserting chunks that this function searches
• Implementation:FlowiseAI_Flowise_GetVectorStoreProviders -- Fetching provider definitions for query configuration
• Implementation:FlowiseAI_Flowise_GetFileChunks -- Browsing raw chunks for comparison with query results