Skip to content

Releases: antonellof/vectX

v0.2.8

19 Apr 17:48
@github-actions github-actions
v0.2.8
3c0560d

Choose a tag to compare

View all tags
v0.2.8 Latest
Latest

vectX v0.2.8

A maintenance release focused on code quality, removing dead code, and fixing all
compiler and Clippy warnings across the workspace.

What's Changed

Code quality

  • Removed an unused deserialize_vector helper from the REST layer (the
    _optional variant covers all call sites).
  • Annotated Qdrant API-compatibility request structs with #[allow(dead_code)]
    and clear comments so it's obvious which fields are accepted-for-compat-but-
    not-yet-routed (UpdateCollectionRequest, RecoverSnapshotRequest,
    CountRequest, SetPayloadRequest, DeletePayloadRequest,
    ClearPayloadRequest, DeleteVectorsRequest, BatchSearchRequest,
    SearchGroupsRequest, DiscoverRequest, ContextPair,
    DiscoverBatchRequest, FacetRequest, BatchQueryRequest,
    QueryGroupsRequest, VectorConfig).
  • Cleaned up dead code in the storage manager and snapshot importer.
  • Suppressed dead-method warnings on intentionally-retained HnswIndex
    helpers (contains, distance).
  • Removed an unused HashMap import from the integration test.

Clippy fixes (workspace-wide)

  • vectx-core: removed an unnecessary Distance::clone() (it's Copy) and
    switched two manual ceiling-division sites to usize::div_ceil.
  • vectx-storage: dropped an empty if branch and replaced two
    Option::map(..).flatten() calls with Option::and_then.
  • vectx-api: replaced while let Some(_) = iter.next() with for _ in iter,
    collapsed a nested if let / else if, and used Option::map where
    appropriate.

Examples

  • Updated examples/data-chatbot/next-env.d.ts for the Next.js 16 typed-routes
    layout (./.next/dev/types/routes.d.ts).

Compatibility

No public API or wire-format changes. This is a drop-in replacement for v0.2.7.

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Installation

Pre-built binaries

Download the appropriate archive for your platform from the assets below.

From crates.io

cargo install vectx

From source

git clone https://github.com/antonellof/vectX.git
cd vectX
cargo build --release

Quick Start

# Start the server
vectx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Assets 10
Loading

v0.2.7: Client-side embeddings & Data Chatbot

20 Dec 17:32
@antonellof antonellof
v0.2.7
1d3de91

Choose a tag to compare

View all tags
v0.2.7: Client-side embeddings & Data Chatbot

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading

v0.2.6 - The Similarity Contract

20 Dec 15:32
@antonellof antonellof
v0.2.6
ac3b98e

Choose a tag to compare

View all tags
v0.2.6 - The Similarity Contract

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading

v0.2.5 - Schema-Driven Similarity Engine

19 Dec 19:22
@antonellof antonellof
v0.2.5
fa9146d

Choose a tag to compare

View all tags
v0.2.5 - Schema-Driven Similarity Engine

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading

v0.2.4

19 Dec 16:19
@github-actions github-actions
v0.2.4
ca1d17e

Choose a tag to compare

View all tags
v0.2.4

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading

v0.2.3 - Version Tracking & Hybrid Search

19 Dec 14:04
@antonellof antonellof
v0.2.3
630f911

Choose a tag to compare

View all tags
v0.2.3 - Version Tracking & Hybrid Search

What's New

Point Version Tracking

  • Points now track their version number, starting at 0 and incrementing on each update
  • Version is included in all point responses (get, scroll, search, query)

Hybrid Search with RRF Fusion

  • Support for prefetch queries in the /points/query endpoint
  • Reciprocal Rank Fusion (RRF) to combine results from multiple search sources
  • Enables combining dense and sparse vector searches for better relevance

API Improvements

  • Sparse vector format support in queries ({"indices": [...], "values": [...]})
  • All point endpoints now return the version field

Installation

cargo install distx

Or with Docker:

docker run -p 6333:6333 distx:0.2.3
Loading

v0.2.2 - Full Qdrant API Parity

19 Dec 13:48
@antonellof antonellof
v0.2.2
f483c2e

Choose a tag to compare

View all tags
v0.2.2 - Full Qdrant API Parity

What's New

Sparse Vector Support

  • Full support for sparse vectors with indices/values format
  • Sparse-only collections now work correctly

Advanced Filtering

  • Complete filter parsing with must, should, must_not conditions
  • Support for match conditions (value, any, text)
  • Support for range conditions (gt, gte, lt, lte)

New API Endpoints

  • Payload operations: set, overwrite, delete, clear
  • Vector operations: update, delete
  • Field index management: create, delete
  • Collection aliases: create, delete, rename, list
  • Batch update operations
  • Service endpoints: /livez, /readyz, /metrics, /issues
  • Search groups, discover, and facet endpoints
  • Full storage snapshot management

Bug Fixes

  • Fixed scroll filtering for non-existent fields
  • Fixed dimension validation for sparse-only collections

Installation

cargo install distx

Or with Docker:

docker pull ghcr.io/antonellof/distx:0.2.2
Loading

v0.2.1: Multivector Support (ColBERT MaxSim)

19 Dec 06:41
@antonellof antonellof
v0.2.1
43c193c

Choose a tag to compare

View all tags
v0.2.1: Multivector Support (ColBERT MaxSim)

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading

v0.2.0: Docker Support & Qdrant Web UI

18 Dec 20:11
@antonellof antonellof
v0.2.0
14d48d4

Choose a tag to compare

View all tags
v0.2.0: Docker Support & Qdrant Web UI

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading

v0.1.1

18 Dec 10:00
@github-actions github-actions
v0.1.1
ef48482

Choose a tag to compare

View all tags
v0.1.1

DistX v0.1.1

High-performance vector database with Qdrant API compatibility.

Highlights

  • 🚀 6x faster search than Qdrant
  • ⚡ 10x faster inserts than Redis
  • 🎯 0.13ms p50 latency (gRPC)
  • 📦 Single binary (~6MB)

Performance

Protocol vs Qdrant vs Redis
Insert (gRPC) 4.5x faster 10x faster
Search (gRPC) 6.3x faster 5x faster
Search Latency 6.6x lower 5x lower

Features

  • HNSW Indexing with SIMD optimizations (AVX2, SSE, NEON)
  • Qdrant-compatible REST API - drop-in replacement
  • gRPC API for maximum performance
  • BM25 text search with ranking
  • Payload filtering by JSON metadata
  • Redis-style persistence (WAL + snapshots + LMDB)

Installation

Pre-built Binaries

Download the appropriate binary for your platform from the assets below.

From crates.io

cargo install distx

From Source

git clone https://github.com/antonellof/DistX.git
cd DistX
cargo build --release

Quick Start

# Start the server
distx --http-port 6333 --grpc-port 6334

# Create a collection
curl -X PUT http://localhost:6333/collections/test \
  -H "Content-Type: application/json" \
  -d '{"vectors": {"size": 128, "distance": "Cosine"}}'

# Insert vectors
curl -X PUT http://localhost:6333/collections/test/points \
  -H "Content-Type: application/json" \
  -d '{"points": [{"id": 1, "vector": [0.1, 0.2, ...]}]}'

# Search
curl -X POST http://localhost:6333/collections/test/points/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "limit": 10}'

Links

Loading