#9999th of 102 in Backends, databases and hosting
vearch
Distributed vector database with hybrid vector and scalar-filter search
- Stars
- 2.3k
- License
- Apache-2.0
- Last commit
- Jul 2026
- Last release
- Feb 2026
- Language
- Python
2,289 on 21 Apr 20262,334 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Vearch is a distributed vector database for similarity search over embeddings. A Router exposes a REST API (upsert, delete, search, query), a Master manages schema and cluster metadata, and PartitionServers run the Gamma engine, built on faiss, with raft replication. It ships SDKs for Python, Go, Java and Rust, plus integrations for LangChain and LlamaIndex.
Who it is for: Engineers self-hosting a vector store for RAG or image similarity search
Strengths
- Combines vector search with scalar filtering in one query
- Raft-based replication and scale-out across Master, Router and PartitionServer roles
- SDKs for Python, Go, Java and Rust, plus LangChain and LlamaIndex integrations
- Helm chart and Docker Compose profiles for standalone and cluster modes
Weaknesses
- Cluster mode needs several components, which is heavier than an embedded vector store
- README gives no benchmarks beyond a 'millions of objects in milliseconds' claim
- Visual search demo needs a separate image retrieval plugin
Quick answers
Is vearch open source?
Yes. Its code is on GitHub under the Apache-2.0 license.
Also in Backends, databases and hosting
See all 102| Rank | Project | Score |
|---|---|---|
| 1 | bunJavaScript and TypeScript runtime with bundler, test runner and package manager | 93 out of 100 |
| 2 | strapiHeadless CMS that generates REST and GraphQL APIs from content models | 92 out of 100 |
| 3 | traefikHTTP reverse proxy and load balancer that auto-discovers routes from orchestrators | 92 out of 100 |
| #4 | appwriteBackend with auth, databases, storage, functions and an MCP server for agents | 92 out of 100 |
| #5 | medusaNode.js commerce platform with modular products, carts, orders and payments | 92 out of 100 |