Memory Vault
Self-hosted AI memory on Postgres and pgvector, with hybrid search and a knowledge graph
- Pricing
- Open source
- Platforms
- self-hosted, API, CLI
- Runs
- Locally
- Version
- 1.7.0
- Listing updated
- 1 day ago
About Memory Vault
Memory Vault is an open-source MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.
Memory Vault is a memory database you run yourself, aimed at ending the habit of re-explaining your project to an assistant in every new chat. It stores decisions, notes, conversations and project context in one Postgres database with pgvector, and searches them with a blend of semantic similarity and keyword matching merged by reciprocal rank fusion, so exact terms and paraphrases both turn up.
The MCP server gives Claude four tools, recall, remember, forget and status. Beyond that, the project includes a dashboard, a bearer-protected REST API documented at /docs, and a knowledge graph built with spaCy entity recognition rather than an LLM, so there is no per-token cost. A local-LLM chat through LM Studio shows the retrieved chunks behind each answer.
It is MIT licensed, ships as a single Docker image with docker compose, and reached v1.0 on 2026-05-07. The MCP server runs from a container image and needs Postgres connection settings.
Install
Run in a terminal
claude mcp add memory-vault --env DB_HOST='<DB_HOST>' --env DB_PORT='<DB_PORT>' --env DB_NAME='<DB_NAME>' --env DB_USER='<DB_USER>' --env DB_PASSWORD='<DB_PASSWORD>' -- docker run -i --rm -e DB_HOST -e DB_PORT -e DB_NAME -e DB_USER -e DB_PASSWORD ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0Add to claude_desktop_config.json
{
"mcpServers": {
"memory-vault": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"DB_HOST",
"-e",
"DB_PORT",
"-e",
"DB_NAME",
"-e",
"DB_USER",
"-e",
"DB_PASSWORD",
"ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0"
],
"env": {
"DB_HOST": "<DB_HOST>",
"DB_PORT": "<DB_PORT>",
"DB_NAME": "<DB_NAME>",
"DB_USER": "<DB_USER>",
"DB_PASSWORD": "<DB_PASSWORD>"
}
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"memory-vault": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"DB_HOST",
"-e",
"DB_PORT",
"-e",
"DB_NAME",
"-e",
"DB_USER",
"-e",
"DB_PASSWORD",
"ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0"
],
"env": {
"DB_HOST": "<DB_HOST>",
"DB_PORT": "<DB_PORT>",
"DB_NAME": "<DB_NAME>",
"DB_USER": "<DB_USER>",
"DB_PASSWORD": "<DB_PASSWORD>"
}
}
}
}Add to .vscode/mcp.json
{
"servers": {
"memory-vault": {
"type": "stdio",
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"DB_HOST",
"-e",
"DB_PORT",
"-e",
"DB_NAME",
"-e",
"DB_USER",
"-e",
"DB_PASSWORD",
"ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0"
],
"env": {
"DB_HOST": "<DB_HOST>",
"DB_PORT": "<DB_PORT>",
"DB_NAME": "<DB_NAME>",
"DB_USER": "<DB_USER>",
"DB_PASSWORD": "<DB_PASSWORD>"
}
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"memory-vault": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"DB_HOST",
"-e",
"DB_PORT",
"-e",
"DB_NAME",
"-e",
"DB_USER",
"-e",
"DB_PASSWORD",
"ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0"
],
"env": {
"DB_HOST": "<DB_HOST>",
"DB_PORT": "<DB_PORT>",
"DB_NAME": "<DB_NAME>",
"DB_USER": "<DB_USER>",
"DB_PASSWORD": "<DB_PASSWORD>"
}
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"memory-vault": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"DB_HOST",
"-e",
"DB_PORT",
"-e",
"DB_NAME",
"-e",
"DB_USER",
"-e",
"DB_PASSWORD",
"ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0"
],
"env": {
"DB_HOST": "<DB_HOST>",
"DB_PORT": "<DB_PORT>",
"DB_NAME": "<DB_NAME>",
"DB_USER": "<DB_USER>",
"DB_PASSWORD": "<DB_PASSWORD>"
}
}
}
}Configuration
DB_HOSTHostname of the Postgres+pgvector instance Memory Vault connects to (e.g. 'host.docker.internal' when Postgres runs on the host, 'localhost' when running natively, or the service name when both run in compose)RequiredDB_PORTPostgres portRequiredDB_NAMEPostgres database nameRequiredDB_USERPostgres user with read/write access to the memory vault databaseRequiredDB_PASSWORDPostgres password for DB_USERRequired
Tools this server provides (4)
Taken from the project documentation; the running server may differ.
- recall
- Searches stored memories.
- remember
- Stores a new memory.
- forget
- Removes a memory.
- status
- Reports the vault's status.
What it does
- Hybrid search
- Combines pgvector semantic and keyword search with rank fusion.
- Four MCP tools
- recall, remember, forget and status from Claude.
- Knowledge graph
- Extracts entities and relations with spaCy, no LLM.
- Sourced local chat
- LM Studio answers show their source chunks.
- REST API
- Bearer-auth API with OpenAPI docs.
- One-command Docker
- Bundles Postgres, the app and the language model.
Good for
- Let Claude remember project decisions across sessions
- Search your own notes by meaning or exact term
- Build an AI tool on a self-hosted memory API
- Browse how people and topics connect in your notes
Pricing
Open source
MIT licence, self-hosted
Prices change. Check mihaibuilds.com before you buy.
Memory Vault FAQ
Where does the data live?
In your own Postgres database, self-hosted.
What licence is it?
MIT.
How do I run it?
docker compose up -d, then open localhost:8000.
Does the graph cost per token?
No, entity extraction uses spaCy rather than an LLM.
Is Memory Vault free?
Yes. Memory Vault is open source, so you can use and self-host it for free.
How do I add Memory Vault to Claude Code?
Run this command in your terminal: claude mcp add memory-vault --env DB_HOST='<DB_HOST>' --env DB_PORT='<DB_PORT>' --env DB_NAME='<DB_NAME>' --env DB_USER='<DB_USER>' --env DB_PASSWORD='<DB_PASSWORD>' -- docker run -i --rm -e DB_HOST -e DB_PORT -e DB_NAME -e DB_USER -e DB_PASSWORD ghcr.io/mihaibuilds/memory-vault-mcp:1.7.0
What are the best alternatives to Memory Vault?
The closest alternatives to Memory Vault are REA, Codebase Memory, XERJ, Claude Code Ultimate Guide and Nautilus Compass Memory.
Alternatives to Memory Vault

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Claude Code Ultimate Guide
MCP serverMCP server for searching the Claude Code Ultimate Guide from any MCP client
Runs locally- Open source
- 6.5K GitHub stars
- AI models and LLM tooling

Nautilus Compass Memory
MCP serverLocal-first long-term memory layer for AI agents, served over MCP
Not checked yet- 5 tools
- Remote
- Open source
- 1.3K GitHub stars
- Knowledge and memory
Agent Roast
MCP serverOffline review that flags agent changes which weaken tests, assertions or CI
Runs locally- Open source
- 4 Hacker News points
- Developer tools