Which vector or graph database should I use?

Answer a few questions and get three picks from 50 catalogued vector databases, graph stores, search engines and embedding services. Every reason is a catalogued fact with its source. Picks are ranked on fit; GitHub stars only break ties. Archived and shut-down systems are left out, and only one product per vendor is recommended. Where the catalog can't confirm a must-have, the system goes under might fit.

What do you need?
Language
Where should it run?
Must-haves

Top picks

1 Milvus Vector-database infrastructure

Multi-vector columns (10 simultaneous). Native hybrid search (v2.5). CAGRA + Vamana GPU/CPU (v2.6).

  • SDKs: Python, Go, Java, Node.js, C++ (+ RESTful API; C# community/Microsoft, Rust in development)↗
  • Licence: Apache-2.0↗
  • Deployment: Both (self-hosted OSS + Zilliz Cloud managed)↗
  • MCP: yes — official zilliztech/mcp-server-milvus (GitHub)↗
  • Pricing: free (OSS); managed via Zilliz Cloud (free tier, then usage-based)↗
  • Recently released: v3.0.3 (2026-10-14)↗
  • Adoption (tiebreaker): milvus-io/milvus — ~46k stars, last commit 2026-10↗

Alternatives · What it works with

2 Qdrant Vector-database infrastructure

Distribution-Based Score Fusion + RRF. Sparse vectors native; filtering via ANN graph modification.

  • SDKs: Python, JavaScript/TypeScript, Rust, Go, .NET, Java↗
  • Licence: Apache-2.0↗
  • Deployment: Both (self-hosted + managed cloud) — OSS engine; Qdrant Managed Cloud, Hybrid Cloud and Private Cloud↗
  • MCP: yes — official qdrant/mcp-server-qdrant (GitHub, Apache-2.0)↗
  • Pricing: Free tier forever (single node, 1 GB RAM / 4 GB disk); Standard usage-based hourly billing; Premium has minimum spend; Hybrid/Private Cloud via sales↗
  • Recently released: v1.19.2 (2026-10-05)↗
  • Adoption (tiebreaker): qdrant/qdrant — ~34.9k stars, last commit 2026-10↗

Alternatives · What it works with

3 pgvector Vector-database infrastructure

Stores embeddings alongside relational + full-text data. HNSW + IVFFlat ANN indexes. Used as agent conversation memory via LangChain + MCP. Foundation of Supabase AI and many self-hosted RAG stacks.

  • SDKs: Client libs for 35 languages incl. Python, JavaScript/TypeScript, Go, Java, Rust, C#, Ruby, PHP↗
  • Deployment: Self-hosted only — PostgreSQL extension; also offered inside managed Postgres (Supabase, AWS RDS, Azure, Neon, etc.)↗
  • MCP: yes — community Postgres MCP servers with pgvector tools (e.g. MOmayr/mcp-postgres-pgvector)↗
  • Pricing: free (OSS)↗
  • Recently released: v0.8.6 (2026-07-29)↗
  • Adoption (tiebreaker): pgvector/pgvector — ~23.5k stars, last commit 2026-07 or later (0.8.7 unreleased)↗

Alternatives · What it works with

Also consider

23 systems considered · 1 ruled out by your must-haves or because they're archived / shut down.

Compare side by side

CompareSystemSelf-hostManagedOpen sourceMCPGitHub ★Latest releaseLicence
Milvus✓✓✓✓46k2026-10-14Apache-2.0
Qdrant✓✓✓✓35k2026-10-05Apache-2.0
pgvector✓✓—✓24k2026-07-29PostgreSQL
Weaviate✓✓✓✓17k2026-10-21BSD-3-Clause
LanceDB✓✓✓✓11k2026-10-07Apache-2.0
Elasticsearch✓✓✓✓78kAGPL-3.0 / SSPL-1.0 /…
Activeloop Deep Lake✓✓✓?9.3k2026-08-25Apache-2.0
Vespa✓✓✓?7.1kApache-2.0
Algolia—✓—✓proprietary (hosted SaaS)…
Supabase Vector✓✓✓✓108kApache-2.0 (Supabase stack)…
OpenSearch✓?✓✓14kApache-2.0
MongoDB Atlas Vector Search✓✓?✓
Pinecone✓✓?✓
Snowflake Cortex Search—✓?✓
Turbopuffer✓✓??
Chromaacquired✓✓✓✓29k2026-10-13Apache-2.0
Typesense✓✓✓✓27kGPL-3.0
Tantivy✓✓✓?16kMIT
Databricks Vector Search (now Databricks AI Search)—✓?✓
Vald✓✓✓?1.7kApache-2.0
Quickwitacquired✓✓✓?12kApache-2.0
Sajari (now Search.io)acquired????4

✓ confirmed · — no · ? not in the catalog yet. Hover a cell for the catalogued text. Already using some of these? Check your stack.