# Mem0 — Long-Term Memory Layer for AI Agents

> Add persistent memory to AI agents and assistants. Remembers user preferences, context, and past interactions across sessions.

## Install

Save in your project root:

## Quick Use

```bash
pip install mem0ai
```

```python
from mem0 import Memory

memory = Memory()

# Add a memory
memory.add("I prefer dark mode and vim keybindings", user_id="alice")

# Search memories
results = memory.search("What does Alice prefer?", user_id="alice")
print(results)
```

**CLI alternative**:
```bash
npm install -g @mem0/cli
mem0 init
mem0 add "Likes Python and FastAPI" --user-id alice
mem0 search "What does Alice like?" --user-id alice
```

---

## Intro

Mem0 is a memory layer for AI applications. It remembers user preferences, past interactions, and context across sessions — so your AI assistant doesn't start from scratch every conversation. Multi-level memory (user, session, agent state), 26% more accurate than OpenAI's built-in memory, 91% faster, and 90% fewer tokens than full-context approaches. 51,000+ GitHub stars, Y Combinator S24.

---

## Key Features

- **Multi-level memory**: User preferences, session context, agent state
- **Adaptive learning**: Continuously improves from interactions
- **Cross-platform SDKs**: Python (`pip install mem0ai`) and Node.js (`npm install mem0ai`)
- **Self-hosted or managed**: Full control locally, or use the hosted platform
- **LLM-agnostic**: Works with OpenAI, Anthropic, and other providers

## Full Example: Chat with Memory

```python
from openai import OpenAI
from mem0 import Memory

openai_client = OpenAI()
memory = Memory()

def chat_with_memories(message: str, user_id: str = "default_user") -> str:
    # Retrieve relevant memories
    relevant = memory.search(query=message, user_id=user_id, limit=3)
    memories_str = "\
".join(f"- {m['memory']}" for m in relevant["results"])

    # Build prompt with memory context
    system = f"You are a helpful AI. Use these memories:\
{memories_str}"
    messages = [
        {"role": "system", "content": system},
        {"role": "user", "content": message}
    ]

    # Generate response
    response = openai_client.chat.completions.create(
        model="gpt-4.1-nano-2025-04-14",
        messages=messages
    )
    reply = response.choices[0].message.content

    # Save new memories from conversation
    messages.append({"role": "assistant", "content": reply})
    memory.add(messages, user_id=user_id)

    return reply

# Usage
chat_with_memories("I'm a Python developer who loves FastAPI")
chat_with_memories("What framework should I use for my next API?")
# → Mem0 remembers you prefer FastAPI
```

## Use Cases

- **AI Assistants**: Consistent, personalized conversations across sessions
- **Customer Support**: Remember past tickets, preferences, account history
- **Healthcare**: Track patient preferences and interaction history
- **Productivity**: Personal AI that knows your workflows and tools
- **Gaming**: NPCs that remember player choices and adapt

## Integrations

- ChatGPT browser extension (adds memory to ChatGPT, Perplexity, Claude)
- LangGraph for building memory-enhanced agents
- CrewAI for multi-agent systems with shared memory

---



### FAQ

**Q: What is Mem0?**
A: Add persistent memory to AI agents and assistants. Remembers user preferences, context, and past interactions across sessions.

**Q: How do I install Mem0?**
A: Check the Quick Use section above for step-by-step installation instructions. Most assets can be set up in under 2 minutes.

## Source & Thanks

> Created by [Taranjeet Singh](https://github.com/taranjeetsingh) / [mem0ai](https://github.com/mem0ai). Licensed under Apache 2.0.
> [mem0](https://github.com/mem0ai/mem0) — ⭐ 51,300+
> Docs: [docs.mem0.ai](https://docs.mem0.ai)

Thanks to the Mem0 team (YC S24) for building the most popular open-source memory layer for AI.

---
Source: https://tokrepo.com/en/workflows/952ce6d3-b141-4db3-83d4-923a7268aef8
Author: Agent Toolkit