JavaScript/TypeScript Client

Installation

Install via npm:

npm install @qdrant/js-client-rest

Or with yarn:

yarn add @qdrant/js-client-rest

Quick Start

import { QdrantClient } from '@qdrant/js-client-rest';

const client = new QdrantClient({ url: 'http://localhost:6333' });

// Create collection
await client.createCollection('my_collection', {
  vectors: {
    size: 384,
    distance: 'Cosine',
  },
});

// Upsert points
await client.upsert('my_collection', {
  points: [
    {
      id: 1,
      vector: [0.05, 0.61, 0.76, 0.74],
      payload: { city: 'Berlin', country: 'Germany' },
    },
    {
      id: 2,
      vector: [0.19, 0.81, 0.75, 0.11],
      payload: { city: 'London', country: 'UK' },
    },
  ],
});

// Search
const results = await client.search('my_collection', {
  vector: [0.2, 0.1, 0.9, 0.7],
  limit: 3,
});

console.log(results);

Repository

GitHub Repository

Official JavaScript client: qdrant/qdrant-js

Client Initialization

Basic Connection

import { QdrantClient } from '@qdrant/js-client-rest';

// Connect to local instance
const client = new QdrantClient({
  url: 'http://localhost:6333',
});

// Connect to Qdrant Cloud
const cloudClient = new QdrantClient({
  url: 'https://your-cluster.cloud.qdrant.io',
  apiKey: 'your-api-key',
});

With Custom Options

const client = new QdrantClient({
  url: 'http://localhost:6333',
  apiKey: 'your-api-key',
  timeout: 30000, // Request timeout in ms
  headers: {
    'Custom-Header': 'value',
  },
});

TypeScript Support

The client has full TypeScript definitions:

import { QdrantClient, Schemas } from '@qdrant/js-client-rest';

type PointPayload = {
  title: string;
  category: string;
  views: number;
  tags: string[];
};

const client = new QdrantClient({ url: 'http://localhost:6333' });

// Type-safe point structure
const point: Schemas['PointStruct'] = {
  id: 1,
  vector: [0.1, 0.2, 0.3],
  payload: {
    title: 'Document 1',
    category: 'technology',
    views: 1500,
    tags: ['ai', 'ml'],
  } as PointPayload,
};

await client.upsert('my_collection', {
  points: [point],
});

// Type-safe search results
const results = await client.search('my_collection', {
  vector: [0.2, 0.1, 0.9],
  limit: 10,
});

results.forEach((result) => {
  console.log(result.id);
  console.log(result.score);
  const payload = result.payload as PointPayload;
  console.log(payload.title);
});

Collection Management

Create Collection

// Single vector
await client.createCollection('my_collection', {
  vectors: {
    size: 768,
    distance: 'Cosine',
    on_disk: false,
  },
});

// Multiple named vectors
await client.createCollection('multi_vector', {
  vectors: {
    text: {
      size: 768,
      distance: 'Cosine',
    },
    image: {
      size: 512,
      distance: 'Euclid',
    },
  },
});

Create with HNSW Config

await client.createCollection('optimized', {
  vectors: {
    size: 384,
    distance: 'Cosine',
    hnsw_config: {
      m: 16,
      ef_construct: 100,
      full_scan_threshold: 10000,
    },
  },
  optimizers_config: {
    default_segment_number: 2,
    indexing_threshold: 50000,
  },
  replication_factor: 2,
  write_consistency_factor: 1,
});

Get Collection Info

const info = await client.getCollection('my_collection');

console.log(`Status: ${info.status}`);
console.log(`Vectors count: ${info.vectors_count}`);
console.log(`Points count: ${info.points_count}`);
console.log(`Segments: ${info.segments_count}`);

List Collections

const collections = await client.getCollections();

collections.collections.forEach((collection) => {
  console.log(collection.name);
});

Update Collection

await client.updateCollection('my_collection', {
  optimizers_config: {
    indexing_threshold: 50000,
    memmap_threshold: 100000,
  },
});

Delete Collection

await client.deleteCollection('my_collection');

Point Operations

Upsert Points

import { v4 as uuidv4 } from 'uuid';

await client.upsert('my_collection', {
  points: [
    {
      id: uuidv4(), // UUID string
      vector: [0.05, 0.61, 0.76, 0.74],
      payload: {
        title: 'Document 1',
        category: 'technology',
        views: 1500,
        tags: ['ai', 'ml', 'vector-db'],
      },
    },
    {
      id: 2, // Integer ID
      vector: [0.19, 0.81, 0.75, 0.11],
      payload: { title: 'Document 2' },
    },
  ],
  wait: true, // Wait for operation to complete
});

Batch Upsert

const batchSize = 100;
const totalPoints = 10000;

for (let i = 0; i < totalPoints; i += batchSize) {
  const points = Array.from({ length: batchSize }, (_, j) => ({
    id: i + j,
    vector: Array.from({ length: 384 }, () => Math.random()),
    payload: { index: i + j },
  }));

  await client.upsert('my_collection', {
    points,
    wait: false, // Don't wait for faster throughput
  });
}

Get Points

const points = await client.retrieve('my_collection', {
  ids: [1, 2, 3],
  with_payload: true,
  with_vector: true,
});

points.forEach((point) => {
  console.log(`ID: ${point.id}`);
  console.log(`Vector: ${point.vector}`);
  console.log(`Payload:`, point.payload);
});

Delete Points

// Delete by IDs
await client.delete('my_collection', {
  points: [1, 2, 3],
});

// Delete by filter
await client.delete('my_collection', {
  filter: {
    must: [
      {
        key: 'category',
        match: { value: 'outdated' },
      },
    ],
  },
});

Search Operations

Basic Search

const results = await client.search('my_collection', {
  vector: [0.2, 0.1, 0.9, 0.7],
  limit: 10,
  with_payload: true,
  with_vector: false,
});

results.forEach((result) => {
  console.log(result.id, result.score);
  console.log(result.payload);
});

Search with Filtering

const results = await client.search('my_collection', {
  vector: [0.2, 0.1, 0.9, 0.7],
  filter: {
    must: [
      {
        key: 'category',
        match: { value: 'technology' },
      },
      {
        key: 'views',
        range: {
          gte: 1000,
          lt: 10000,
        },
      },
    ],
  },
  limit: 5,
  score_threshold: 0.7,
});

Search with Parameters

const results = await client.search('my_collection', {
  vector: [0.2, 0.1, 0.9, 0.7],
  limit: 20,
  params: {
    hnsw_ef: 128, // Higher ef for better recall
    exact: false, // Use approximate search
    quantization: {
      ignore: false,
      rescore: true,
      oversampling: 2.0,
    },
  },
});

Batch Search

const results = await client.searchBatch('my_collection', {
  searches: [
    {
      vector: [0.2, 0.1, 0.9, 0.7],
      limit: 5,
      filter: {
        must: [{ key: 'category', match: { value: 'tech' } }],
      },
    },
    {
      vector: [0.5, 0.3, 0.2, 0.8],
      limit: 10,
    },
  ],
});

results.forEach((batch, index) => {
  console.log(`Results for query ${index}:`);
  batch.forEach((point) => console.log(point.id, point.score));
});

Recommendations

Basic Recommendations

const results = await client.recommend('my_collection', {
  positive: [1, 2, 3], // Point IDs to use as positive examples
  negative: [10], // Point IDs to avoid
  limit: 10,
});

const results = await client.recommend('my_collection', {
  positive: [1, 2],
  negative: [10],
  strategy: 'best_score', // or 'average_vector'
  limit: 5,
  filter: {
    must: [{ key: 'category', match: { value: 'tech' } }],
  },
});

Batch Recommendations

const results = await client.recommendBatch('my_collection', {
  searches: [
    {
      positive: [1, 2],
      negative: [10],
      limit: 5,
    },
    {
      positive: [3, 4, 5],
      limit: 10,
    },
  ],
});

Query API (Universal)

// Hybrid search with fusion
const results = await client.query('my_collection', {
  prefetch: [
    {
      query: [0.1, 0.2, 0.3], // Dense vector
      using: 'dense',
      limit: 100,
    },
    {
      query: { indices: [1, 5, 10], values: [0.5, 0.3, 0.2] }, // Sparse vector
      using: 'sparse',
      limit: 100,
    },
  ],
  query: { fusion: 'rrf' }, // Reciprocal Rank Fusion
  limit: 10,
});

Payload Operations

Set Payload

await client.setPayload('my_collection', {
  payload: {
    new_field: 'value',
    updated_at: '2024-01-01',
  },
  points: [1, 2, 3],
});

Overwrite Payload

await client.overwritePayload('my_collection', {
  payload: { only_field: 'value' },
  points: [1, 2],
});

Delete Payload Keys

await client.deletePayload('my_collection', {
  keys: ['old_field', 'deprecated'],
  points: [1, 2, 3],
});

Clear Payload

await client.clearPayload('my_collection', {
  points: [1, 2, 3],
});

Scroll (Pagination)

let offset: string | number | null = null;
const allPoints = [];

while (true) {
  const result = await client.scroll('my_collection', {
    limit: 100,
    offset: offset || undefined,
    with_payload: true,
    filter: {
      must: [{ key: 'category', match: { value: 'tech' } }],
    },
  });

  allPoints.push(...result.points);
  offset = result.next_page_offset;

  if (!offset) break;
}

console.log(`Total points: ${allPoints.length}`);

Payload Indexing

// Create keyword index
await client.createPayloadIndex('my_collection', {
  field_name: 'category',
  field_schema: 'keyword',
});

// Create text index for full-text search
await client.createPayloadIndex('my_collection', {
  field_name: 'description',
  field_schema: {
    type: 'text',
    tokenizer: 'word',
    lowercase: true,
    min_token_len: 2,
    max_token_len: 20,
  },
});

// Delete index
await client.deletePayloadIndex('my_collection', 'category');

Snapshots

Create Snapshot

const snapshot = await client.createSnapshot('my_collection');
console.log(`Snapshot name: ${snapshot.name}`);

List Snapshots

const snapshots = await client.listSnapshots('my_collection');
snapshots.forEach((snapshot) => {
  console.log(snapshot.name);
});

Recover from Snapshot

await client.recoverSnapshot('my_collection', {
  location: 'http://localhost:6333/snapshots/my_collection/snapshot.snapshot',
});

Error Handling

import { QdrantClient } from '@qdrant/js-client-rest';

const client = new QdrantClient({ url: 'http://localhost:6333' });

try {
  await client.getCollection('nonexistent');
} catch (error: any) {
  if (error.status === 404) {
    console.log('Collection not found');
  } else {
    console.error(`Error: ${error.status} - ${error.message}`);
  }
}

Usage with Node.js

const { QdrantClient } = require('@qdrant/js-client-rest');

const client = new QdrantClient({ url: 'http://localhost:6333' });

async function main() {
  // Create collection
  await client.createCollection('my_collection', {
    vectors: {
      size: 384,
      distance: 'Cosine',
    },
  });

  // Upsert points
  await client.upsert('my_collection', {
    points: [
      {
        id: 1,
        vector: [0.05, 0.61, 0.76, 0.74],
        payload: { name: 'Document 1' },
      },
    ],
  });

  // Search
  const results = await client.search('my_collection', {
    vector: [0.2, 0.1, 0.9, 0.7],
    limit: 5,
  });

  console.log(results);
}

main().catch(console.error);

Usage with Browser

<!DOCTYPE html>
<html>
  <head>
    <script type="module">
      import { QdrantClient } from 'https://cdn.jsdelivr.net/npm/@qdrant/js-client-rest/+esm';

      const client = new QdrantClient({
        url: 'http://localhost:6333',
      });

      async function search() {
        const results = await client.search('my_collection', {
          vector: [0.2, 0.1, 0.9, 0.7],
          limit: 5,
        });
        console.log(results);
      }

      search();
    </script>
  </head>
  <body>
    <h1>Qdrant Browser Example</h1>
  </body>
</html>

Advanced Filtering

// Complex filter with nested conditions
const results = await client.search('my_collection', {
  vector: [0.2, 0.1, 0.9],
  filter: {
    must: [
      {
        key: 'category',
        match: { value: 'technology' },
      },
    ],
    should: [
      {
        key: 'views',
        range: { gte: 10000 },
      },
      {
        key: 'featured',
        match: { value: true },
      },
    ],
    must_not: [
      {
        key: 'status',
        match: { value: 'archived' },
      },
    ],
  },
  limit: 10,
});

Best Practices

Performance tips:

Use batch operations for multiple points
Set wait: false for faster upserts
Create payload indexes before filtering
Use appropriate batch sizes (100-1000 points)

Common issues:

Always handle errors with try-catch
Ensure vector dimensions match collection config
Use UUID strings or integers for point IDs
Handle pagination when scrolling large datasets

Next Steps

Python Client

Python client documentation

Rust Client

Rust client library

​Installation

​Quick Start

​Repository

GitHub Repository

​Client Initialization

​Basic Connection

​With Custom Options

​TypeScript Support

​Collection Management

​Create Collection

​Create with HNSW Config

​Get Collection Info

​List Collections

​Update Collection

​Delete Collection

​Point Operations

​Upsert Points

​Batch Upsert

​Get Points

​Delete Points

​Search Operations

​Basic Search

​Search with Filtering

​Search with Parameters

​Batch Search

​Recommendations

​Basic Recommendations

​Recommend with Strategy

​Batch Recommendations

​Query API (Universal)

​Payload Operations

​Set Payload

​Overwrite Payload

​Delete Payload Keys

​Clear Payload

​Scroll (Pagination)

​Payload Indexing

​Snapshots

​Create Snapshot

​List Snapshots

​Recover from Snapshot

​Error Handling

​Usage with Node.js

​Usage with Browser

​Advanced Filtering

​Best Practices

​Next Steps