mcpbeat Sign in

pgvector MCP Server

local only

pgvector runs on your own machine — the client starts it, so there is no endpoint to ping. 62 installs a week from pypi. Last commit 21 Jul 2026.

Similarity search, hybrid search, and index management for pgvector-backed PostgreSQL tables

Installs per day peak 16 · avg 8 · -11% w/w
a month agotoday
62
Installs / week
pypi · mcp-server-pgvector
0
Stars
0 open issues
21 Jul 2026
Last commit
0 releases in 90 days
MIT
License
Python

Nothing serious here today

Today is the operative word: we check pgvector every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.

Three servers free · no card

Connect this server

This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.

run in your terminal
claude mcp add pgvector -- uvx mcp-server-pgvector
~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "pgvector": {
      "args": [
        "mcp-server-pgvector"
      ],
      "command": "uvx"
    }
  }
}
~/.codex/config.toml
[mcp_servers.pgvector]
command = "uvx"
args = ["mcp-server-pgvector"]
.cursor/mcp.json
{
  "mcpServers": {
    "pgvector": {
      "args": [
        "mcp-server-pgvector"
      ],
      "command": "uvx"
    }
  }
}
.vscode/mcp.json
{
  "mcpServers": {
    "pgvector": {
      "args": [
        "mcp-server-pgvector"
      ],
      "command": "uvx"
    }
  }
}

This one needs environment variables set before it will start: DATABASE_URL (PostgreSQL connection string, e.g. postgresql://user:password@host:5432/dbname), MCP_PGVECTOR_READ_ONLY (Set to 'true' to disable upsert_embedding and create_vector_index). The author declared them in the registry entry; get the values from the project itself.

Alternatives to pgvector

same job, measured the same way
Code Index
by achreftlili

SQLite-backed hybrid (vector + FTS) code search index for Claude Code over MCP.

213 installs/wk local only
Postgres MCP
by pgedge

Enterprise PostgreSQL MCP server with NL queries, hybrid search (pgvector+BM25), and web UI

local only
Memory Vault
by mihaibuilds

Local-first AI memory layer with hybrid search. Postgres + pgvector. Self-hosted, MIT.

local only
Code Context
by infino-ai

Local code search for AI coding agents: hybrid search and SQL aggregation over an in-repo index.

93 installs/wk local only
PostgreSQL MCP Server
by itunified-io

PostgreSQL MCP Server — 27 tools for queries, schema, CRUD, DBA, HA & server management

23 installs/wk local only
Berth
by seayniclabs

Database management — query, schema inspection, migrations for PostgreSQL, MySQL, SQLite.

91 installs/wk local only
MCP DB Connect
by phatngoit

MCP server for Oracle, MSSQL, PostgreSQL, MySQL, SQLite, MongoDB, and Qdrant vector search access.

61 installs/wk local only
Virtual Filesystem
by lu-zhengda

PostgreSQL-backed virtual filesystem for agents: session isolation, shared stores, POSIX ops, RLS.

29 installs/wk local only

pgvector — questions

Answers built from our own checks of this server.

Why is there no uptime for pgvector?
pgvector runs on your own machine over stdio — there is no network address to reach, so uptime cannot be measured for it by anyone. What can be measured is adoption: the pypi package mcp-server-pgvector was installed 62 times last week.
How do I connect pgvector?
Copy the ready config from this page — we generate it for Claude Code, Claude Desktop, Codex, Cursor and VS Code, each with the file path that client actually reads. It runs locally, so the command pulls mcp-server-pgvector straight from pypi; nothing to host, nothing to sign up for.
How many people use pgvector?
The pypi package mcp-server-pgvector was installed 62 times in the last week. Week over week that is -11%. We show installs rather than GitHub stars on purpose: a star is a bookmark, an install is someone actually running it.
Is pgvector open source?
Yes — it is published under the MIT licence, written in Python and 0 stars on GitHub. The source link is on this page, so you can read exactly what it does with your data before you connect it.