Self-hosted vector database

Milvus leads with a score of 91. Updated from public GitHub data.

Full category: Vector databases.

Which one?

1st pickMilvus
Distributed vector database with dense, sparse and hybrid search at scale
Score 91 · 46.7k stars · Apache-2.0
Switch toQdrant
Rust vector database with payload filtering, REST and gRPC
as a strong alternative · Score 85 · 35.6k stars · Apache-2.0
Switch toMeilisearch
Rust search engine with typo tolerance, filtering and hybrid semantic search
to run without a GPU · Score 84 · 59.9k stars · custom license
Self-hosted vector database
RankProjectAdoptionFreshnessMaintenanceAgent-readyScore
1MilvusDistributed vector database with dense, sparse and hybrid search at scale Live demo ↗ (opens in a new tab)46.7k starsApache-2.0last commit Nov 2026known advisory in latest version94100857091 out of 100
2QdrantRust vector database with payload filtering, REST and gRPC Live demo ↗ (opens in a new tab)35.6k starsApache-2.0last commit Nov 202688100824085 out of 100
3MeilisearchRust search engine with typo tolerance, filtering and hybrid semantic search Live demo ↗ (opens in a new tab)59.9k starscustom licenselast commit Nov 202697100743084 out of 100

Momentum, Verified build, Docs and Privacy are not measured yet; their weight goes to the signals shown. A dash means the signal is not scored for that kind of project. Hover a number for its rating in words.

Reviews

191 out of 100

Milvus

Distributed vector database with dense, sparse and hybrid search at scale

Live demo ↗ (opens in a new tab)46.7k starsApache-2.0last commit Nov 2026known advisory in latest version

Milvus is a distributed vector database (Go and C++, LF AI & Data Foundation) that separates compute and storage on Kubernetes, with a Standalone Docker mode and pip-installable Milvus Lite. It offers HNSW, IVF, FLAT, SCANN and DiskANN indexes, GPU CAGRA, sparse BM25 and learned-sparse vectors for hybrid search, metadata filtering, multi-tenancy, hot/cold storage, auth, TLS and RBAC.

Strengths

  • Index types HNSW, IVF, FLAT, SCANN, DiskANN plus GPU CAGRA
  • Dense, sparse (BM25, SPLADE, BGE-M3) and hybrid search in one collection
  • Multi-tenancy at database, collection, partition or partition-key level
  • Mandatory auth, TLS and RBAC; Milvus Lite via pip for local dev

Weaknesses

  • Distributed mode is Kubernetes-native with several microservices to operate
  • No port, RAM or Docker command in the README; install lives in docs
  • Zilliz is the major contributor and promotes its managed cloud
  • Source build needs Go 1.21+, CMake, GCC 11+ and Python 3.8 to 3.11
  • GPU optional
  • Compose
  • Compose runs Milvus, MinIO
  • Models: any embedding model or service; pymilvus[model] wraps embedding and reranking models
285 out of 100

Qdrant

Rust vector database with payload filtering, REST and gRPC

Live demo ↗ (opens in a new tab)35.6k starsApache-2.0last commit Nov 2026

Qdrant is a Rust vector database exposing REST (OpenAPI 3.0) and gRPC on port 6333 for storing points (vectors plus JSON payload) and searching with dense, sparse and multivector (ColBERT) embeddings, rich payload filters and hybrid fusion (RRF, DBSF). It adds quantization, on-disk storage, sharding and replication, multitenancy, GPU-accelerated indexing and a web UI. Qdrant Edge runs the same engine embedded in-process.

Strengths

  • Dense, sparse and multivector (ColBERT) search with RRF and DBSF fusion
  • Quantization cuts RAM up to 97 percent; on-disk storage and io_uring
  • REST with OpenAPI 3.0 spec plus gRPC; six official clients
  • Sharding and replication with zero-downtime collection resize

Weaknesses

  • Default docker run has no auth and binds all interfaces
  • GPU acceleration covers indexing only; search runs on CPU
  • Qdrant Edge embedded mode is Python and Rust only
  • Sharding and tenant isolation require upfront design
  • GPU optional
  • Docker
  • Models: any embedding model; dense, sparse and late-interaction (ColBERT) vectors
  • port 6333
384 out of 100

Meilisearch

Rust search engine with typo tolerance, filtering and hybrid semantic search

Live demo ↗ (opens in a new tab)59.9k starscustom licenselast commit Nov 2026

Meilisearch is a search engine with a RESTful API for full-text, semantic and hybrid search, including typo tolerance, facets, sorting, synonyms and geosearch. It runs as a single binary or Docker image on port 7700, and ships an experimental MCP server at /mcp plus a LangChain vector store integration. The Community Edition is MIT licensed; sharding is in the Enterprise Edition.

Strengths

  • Single binary or Docker image; quick start is one command on port 7700
  • Typo tolerance, facets, filters, sorting, synonyms and geosearch built in
  • Hybrid full-text plus semantic search through one REST API
  • SDKs for JavaScript, Python, PHP, Ruby, Go and Rust, plus a React integration

Weaknesses

  • Sharding and other advanced features need the Enterprise Edition and a commercial agreement
  • README recommends Meilisearch Cloud over self-hosting
  • MCP server at /mcp is marked experimental
  • no GPU
  • Docker
  • Models: embedding models for semantic search (providers unknown)
  • port 7700