@kb-labs/adapters-qdrant

Part of KB Labs ecosystem. Works exclusively within KB Labs platform.

High-performance vector database adapter for semantic search and RAG applications using Qdrant.

Overview

| Property | Value | |----------|-------| | Implements | IVectorStore | | Type | core | | Requires | None | | Category | Database / AI |

Features

• Vector Search - Fast nearest neighbor search
• Hybrid Search - Combine dense and sparse vectors
• Filtering - Payload-based filtering with any query
• Batch Operations - Efficient bulk upsert and delete
• Scalable - Handles millions of vectors

Installation

pnpm add @kb-labs/adapters-qdrant

Configuration

Add to your kb.config.json:

{
  "platform": {
    "adapters": {
      "vectorStore": "@kb-labs/adapters-qdrant"
    },
    "adapterOptions": {
      "vectorStore": {
        "url": "http://localhost:6333",
        "collectionName": "kb-vectors",
        "dimension": 1536
      }
    }
  }
}

Options

| Option | Type | Default | Description | |--------|------|---------|-------------| | url | string | - | Qdrant server URL | | apiKey | string | - | API key for authentication (optional) | | collectionName | string | "kb-vectors" | Collection name | | dimension | number | 1536 | Vector dimension (1536 for OpenAI) | | timeout | number | 30000 | Request timeout in ms |

Usage

Via Platform (Recommended)

import { usePlatform } from '@kb-labs/sdk';

const platform = usePlatform();

// Upsert vectors
await platform.vectorStore.upsert([
  {
    id: 'doc-1',
    vector: [0.1, 0.2, ...], // 1536 dimensions
    payload: { title: 'Document 1', category: 'tech' }
  }
]);

// Search
const results = await platform.vectorStore.search({
  vector: queryVector,
  limit: 10,
  filter: { category: 'tech' }
});

// Delete
await platform.vectorStore.delete(['doc-1', 'doc-2']);

Standalone (Testing/Development)

import { createAdapter } from '@kb-labs/adapters-qdrant';

const vectorStore = createAdapter({
  url: 'http://localhost:6333',
  collectionName: 'test-vectors',
  dimension: 1536
});

await vectorStore.upsert([{ id: '1', vector: [...], payload: {} }]);

Adapter Manifest

{
  id: 'qdrant-vectorstore',
  name: 'Qdrant Vector Store',
  version: '1.0.0',
  implements: 'IVectorStore',
  capabilities: {
    search: true,
    batch: true,
    custom: {
      hybridSearch: true,
      filtering: true,
    },
  },
}

FAQ

Use Docker:

docker run -p 6333:6333 qdrant/qdrant

• OpenAI text-embedding-3-small: 1536
• OpenAI text-embedding-3-large: 3072
• Cohere: 1024

Enable hybrid search in your query:

const results = await vectorStore.search({
  vector: denseVector,
  sparseVector: sparseVector,
  limit: 10
});

Related Adapters

| Adapter | Use Case | |---------|----------| | @kb-labs/adapters-openai | Generate embeddings for vectors | | @kb-labs/adapters-mongodb | Document storage alongside vectors |

License

KB Public License v1.1 - KB Labs Team