← back to the directory
Library / SDK Retrieval & Memory

VectorChord

1Faster 2disk-friendly 3vector 4search 5inside 6Postgres

the six Ws · specification

W1 Who

Postgres users who need faster, cheaper vector search than pgvector at larger scale.

W2 What

VectorChord is a Postgres extension, the successor to pgvecto.rs, providing scalable, disk-friendly vector similarity search with RaBitQ compression and pgvector-compatible syntax.

W3 Where

Self-hosted, installed as a Postgres extension via TensorChord's Docker images or PGXN.

W4 When

Actively maintained by TensorChord as of 2026, succeeding the now-superseded pgvecto.rs project.

W5 Why

It exists to give Postgres users higher insert throughput and query speed than pgvector's HNSW without leaving the database.

W6 With

Requires PostgreSQL and is compatible with existing pgvector data types and queries.

W7 Watch

Open source, dual-licensed AGPLv3 or Elastic License v2; self-hosted; data stays inside the operator's own Postgres instance.

Postgres extensiondisk-friendly ANNRaBitQ compressionpgvector-compatible

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.