Vector Stores

npm install @langchain/community

# Examples
npm install chromadb              # Chroma
npm install hnswlib-node          # HNSWlib
npm install @zilliz/milvus2-sdk-node  # Milvus
npm install pg                    # PGVector

import { Chroma } from "@langchain/community/vectorstores/chroma";
import { OpenAIEmbeddings } from "@langchain/openai";

// Create from documents
const vectorStore = await Chroma.fromDocuments(
  docs,
  new OpenAIEmbeddings(),
  {
    collectionName: "my-collection",
    url: "http://localhost:8000", // Chroma server URL
  }
);

// Search
const results = await vectorStore.similaritySearch("query", 4);

npm install @langchain/pinecone

npm install @langchain/weaviate

npm install @langchain/qdrant

import { FaissStore } from "@langchain/community/vectorstores/faiss";
import { OpenAIEmbeddings } from "@langchain/openai";

// Create and save
const vectorStore = await FaissStore.fromDocuments(
  docs,
  new OpenAIEmbeddings()
);
await vectorStore.save("faiss_index");

// Load and search
const loadedStore = await FaissStore.load(
  "faiss_index",
  new OpenAIEmbeddings()
);
const results = await loadedStore.similaritySearch("query", 4);

import { HNSWLib } from "@langchain/community/vectorstores/hnswlib";
import { OpenAIEmbeddings } from "@langchain/openai";

const vectorStore = await HNSWLib.fromDocuments(
  docs,
  new OpenAIEmbeddings()
);

await vectorStore.save("hnswlib_index");

const loadedStore = await HNSWLib.load(
  "hnswlib_index",
  new OpenAIEmbeddings()
);

import { PGVectorStore } from "@langchain/community/vectorstores/pgvector";
import { OpenAIEmbeddings } from "@langchain/openai";
import { PoolConfig } from "pg";

const config: PoolConfig = {
  host: "localhost",
  port: 5432,
  database: "vectordb",
  user: "user",
  password: "password",
};

const vectorStore = await PGVectorStore.initialize(
  new OpenAIEmbeddings(),
  {
    postgresConnectionOptions: config,
    tableName: "documents",
    columns: {
      idColumnName: "id",
      vectorColumnName: "vector",
      contentColumnName: "content",
      metadataColumnName: "metadata",
    },
  }
);

await vectorStore.addDocuments(docs);

import { SupabaseVectorStore } from "@langchain/community/vectorstores/supabase";
import { OpenAIEmbeddings } from "@langchain/openai";
import { createClient } from "@supabase/supabase-js";

const client = createClient(
  process.env.SUPABASE_URL,
  process.env.SUPABASE_PRIVATE_KEY
);

const vectorStore = await SupabaseVectorStore.fromDocuments(
  docs,
  new OpenAIEmbeddings(),
  {
    client,
    tableName: "documents",
    queryName: "match_documents",
  }
);

npm install @langchain/mongodb

import { AzureAISearchVectorStore } from "@langchain/community/vectorstores/azure_aisearch";
import { OpenAIEmbeddings } from "@langchain/openai";

const vectorStore = new AzureAISearchVectorStore(new OpenAIEmbeddings(), {
  endpoint: process.env.AZURE_AI_SEARCH_ENDPOINT,
  key: process.env.AZURE_AI_SEARCH_KEY,
  indexName: "my-index",
});

// From documents
const vectorStore = await VectorStore.fromDocuments(
  docs,
  embeddings,
  config
);

// From texts
const vectorStore = await VectorStore.fromTexts(
  ["text1", "text2"],
  [{ source: "1" }, { source: "2" }],
  embeddings,
  config
);

// From existing store
const vectorStore = new VectorStore(embeddings, config);

// Add documents
await vectorStore.addDocuments(docs);

// Add with IDs
const ids = await vectorStore.addDocuments(docs);

// Similarity search
const results = await vectorStore.similaritySearch("query", k);

// Similarity search with scores
const resultsWithScores = await vectorStore.similaritySearchWithScore(
  "query",
  k
);

// Max marginal relevance search (diversity)
const diverseResults = await vectorStore.maxMarginalRelevanceSearch(
  "query",
  { k: 4, fetchK: 20 }
);

// Filter by metadata
const results = await vectorStore.similaritySearch(
  "query",
  4,
  { source: "wikipedia" }
);

// Delete by IDs
await vectorStore.delete({ ids: ["id1", "id2"] });

// Delete by filter (if supported)
await vectorStore.delete({ filter: { source: "old" } });

const retriever = vectorStore.asRetriever({
  k: 4,
  searchType: "similarity", // or "mmr"
});

const docs = await retriever.invoke("query");