---
title: Introduction
description: 'Welcome to the Mem0 Memory System - A comprehensive memory layer for AI agents'
---

<img
  className="block dark:hidden"
  src="/images/hero-light.svg"
  alt="Hero Light"
/>
<img
  className="hidden dark:block"
  src="/images/hero-dark.svg"
  alt="Hero Dark"
/>

## What is Mem0 Memory System?

The Mem0 Memory System is a comprehensive, self-hosted memory layer designed for AI agents and applications. Built on top of the open-source mem0 framework, it provides persistent, intelligent memory capabilities that enhance AI interactions through contextual understanding and knowledge retention.

<CardGroup cols={2}>
  <Card
    title="Local-First Architecture"
    icon="server"
    href="/essentials/architecture"
  >
    Complete local deployment with Ollama, Neo4j, and Supabase for maximum privacy and control
  </Card>
  <Card
    title="Multi-Provider Support"
    icon="plug"
    href="/llm/configuration"
  >
    Seamlessly switch between OpenAI, Ollama, and other LLM providers
  </Card>
  <Card
    title="Graph Memory"
    icon="project-diagram"
    href="/database/neo4j"
  >
    Advanced relationship mapping with Neo4j for contextual memory connections
  </Card>
  <Card
    title="MCP Integration"
    icon="link"
    href="/guides/mcp-integration"
  >
    Model Context Protocol server for Claude Code and other AI tools
  </Card>
</CardGroup>

## Key Features

<AccordionGroup>
  <Accordion title="Vector Memory Storage">
    High-performance vector search using Supabase with pgvector for semantic memory retrieval and similarity matching.
  </Accordion>

  <Accordion title="Graph Relationships">
    Neo4j-powered knowledge graph for complex entity relationships and contextual memory connections.
  </Accordion>

  <Accordion title="Local LLM Support">
    Full Ollama integration with 20+ local models including Llama, Qwen, and specialized embedding models.
  </Accordion>

  <Accordion title="API-First Design">
    RESTful API with comprehensive memory operations, authentication, and rate limiting.
  </Accordion>

  <Accordion title="Self-Hosted Privacy">
    Complete local deployment ensuring your data never leaves your infrastructure.
  </Accordion>
</AccordionGroup>

## Architecture Overview

The system consists of several key components working together:

```mermaid
graph TB
    A[AI Applications] --> B[MCP Server]
    B --> C[Memory API]
    C --> D[Mem0 Core]
    D --> E[Vector Store - Supabase]
    D --> F[Graph Store - Neo4j]
    D --> G[LLM Provider]
    G --> H[Ollama Local]
    G --> I[OpenAI/Remote]
```

## Current Status: Phase 1 Complete ✅

<Note>
  **Foundation Ready**: All core infrastructure components are operational and tested.
</Note>

| Component | Status | Description |
|-----------|--------|-------------|
| **Neo4j** | ✅ Ready | Graph database running on localhost:7474 |
| **Supabase** | ✅ Ready | Self-hosted database with pgvector on localhost:8000 |
| **Ollama** | ✅ Ready | 21+ local models available on localhost:11434 |
| **Mem0 Core** | ✅ Ready | Memory management system v0.1.115 |

## Getting Started

<CardGroup cols={1}>
  <Card
    title="Quick Start Guide"
    icon="rocket"
    href="/quickstart"
  >
    Get your memory system running in under 5 minutes
  </Card>
</CardGroup>

Ready to enhance your AI applications with persistent, intelligent memory? Let's get started!