vectX

A fast, in-memory vector database with 100% Qdrant API compatibility.

vectX is a drop-in replacement for Qdrant, providing high-performance vector similarity search with full API compatibility.

Quick Start

# Run vectX
docker run -p 6333:6333 antonellofratepietro/vectx:latest

# Create collection
curl -X PUT localhost:6333/collections/products -d '{
  "vectors": {"size": 1536, "distance": "Cosine"}
}'

# Insert points
curl -X PUT localhost:6333/collections/products/points -d '{
  "points": [{
    "id": 1,
    "vector": [0.1, 0.2, 0.3, ...],
    "payload": {"name": "iPhone", "price": 999}
  }]
}'

# Search
curl -X POST localhost:6333/collections/products/points/search -d '{
  "vector": [0.1, 0.2, 0.3, ...],
  "limit": 10
}'

Key Features

Feature	Description
100% Qdrant Compatible	Use existing Qdrant clients, drop-in replacement
High Performance	Fast in-memory vector search with HNSW indexing
Full Vector DB	Search, filter, facets, scroll, recommendations
Persistence	WAL and snapshot support for data durability
REST & gRPC APIs	Full Qdrant API compatibility
Multi-vector Support	Named vectors and sparse vectors

REST API

Full Qdrant API compatibility:

Endpoint	Description
`PUT /collections/{name}`	Create collection
`POST /collections/{name}/points/search`	Vector similarity search
`POST /collections/{name}/points/scroll`	Filter and browse with pagination
`POST /collections/{name}/points/recommend`	Recommendations from examples
`POST /collections/{name}/facet`	Aggregated counts by field
`PUT /collections/{name}/points`	Upsert points
`GET /collections/{name}/points`	Get points by IDs
`DELETE /collections/{name}/points`	Delete points

Examples

Example applications demonstrating vectX usage:

Data Chatbot — Next.js demo with CSV upload and vector search
Fastest RAG Stack — RAG application with Streamlit UI

Installation

# Docker
docker run -p 6333:6333 antonellofratepietro/vectx:latest

# Cargo
cargo install vectx

# Source
git clone https://github.com/antonellof/vectX.git && cd vectX && cargo run --release

Use with Qdrant Clients

import { QdrantClient } from '@qdrant/js-client-rest';
const client = new QdrantClient({ url: 'http://localhost:6333' });

// All standard Qdrant operations work
await client.search('products', { vector: [...], limit: 10 });

Documentation

Quick Start — Get running in 5 minutes
API Reference — REST and gRPC endpoints
Architecture — System design and internals
Docker Guide — Container deployment
Persistence — WAL, snapshots, storage
Performance — Benchmarks and tuning

License

MIT OR Apache-2.0

vectx-core 0.2.7