# Mem0 Integration (/examples/integrations/memory/mem0-integration)



Demonstrates using Mem0 as an external memory service for an Agno agent.

```python title="mem0_integration.py"
"""
Mem0 Integration
================

Demonstrates using Mem0 as an external memory service for an Agno agent.
"""

from agno.agent import Agent, RunOutput
from agno.models.openai import OpenAIChat
from agno.utils.pprint import pprint_run_response

try:
    from mem0 import MemoryClient
except ImportError:
    raise ImportError(
        "mem0 is not installed. Please install it using `uv pip install mem0ai`."
    )


# ---------------------------------------------------------------------------
# Setup
# ---------------------------------------------------------------------------
client = MemoryClient()

user_id = "agno"
messages = [
    {"role": "user", "content": "My name is John Billings."},
    {"role": "user", "content": "I live in NYC."},
    {"role": "user", "content": "I'm going to a concert tomorrow."},
]

# Comment out the following line after running the script once
client.add(messages, user_id=user_id)


# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------
agent = Agent(
    model=OpenAIChat(),
    dependencies={"memory": client.get_all(user_id=user_id)},
    add_dependencies_to_context=True,
)


# ---------------------------------------------------------------------------
# Run Example
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    run: RunOutput = agent.run("What do you know about me?")
    pprint_run_response(run)

    input = [{"role": i.role, "content": str(i.content)} for i in (run.messages or [])]
    client.add(messages, user_id=user_id)
```

## Current Mem0 API [#current-mem0-api]

Before running the copied recipe, replace its dependency lookup with the current filtered, paginated API:

```python
agent = Agent(
    model=OpenAIChat(),
    dependencies={
        "memory": client.get_all(filters={"user_id": user_id}, page=1, page_size=50)["results"]
    },
    add_dependencies_to_context=True,
)
```

This loads one page of memories when the agent is constructed. Fetch subsequent pages if needed; refresh the dependency on later runs if the stored memories change. Mem0 processes writes asynchronously: wait for the seed write's event to finish before fetching its results. See the [Mem0 migration guide](https://docs.mem0.ai/migration/platform-v2-to-v3) for the event-status endpoint and pagination contract.

The final `client.add(messages, ...)` in the source repeats the seed data. To save this run's conversation instead, replace that final call with:

```python
conversation = [
    {"role": message.role, "content": str(message.content)}
    for message in (run.messages or [])
    if message.role in {"user", "assistant"} and message.content
]
client.add(conversation, user_id=user_id)
```

Use a separate user ID for each user whose memories you store.

## Run the Example [#run-the-example]

<Steps>
    <Step title="Set up your virtual environment">
      <CodeBlockTabs defaultValue="Mac">
        <CodeBlockTabsList>
          <CodeBlockTabsTrigger value="Mac">
            Mac
          </CodeBlockTabsTrigger>

          <CodeBlockTabsTrigger value="Windows">
            Windows
          </CodeBlockTabsTrigger>
        </CodeBlockTabsList>

        <CodeBlockTab value="Mac">
          ```bash
          uv venv --python 3.12
          source .venv/bin/activate
          ```
        </CodeBlockTab>

        <CodeBlockTab value="Windows">
          ```bash
          uv venv --python 3.12
          .venv\Scripts\activate
          ```
        </CodeBlockTab>
      </CodeBlockTabs>
    </Step>

  <Step title="Install dependencies">
    ```bash
    uv pip install -U agno mem0ai openai
    ```
  </Step>

  <Step title="Export your API keys">
    <CodeBlockTabs defaultValue="Mac/Linux">
      <CodeBlockTabsList>
        <CodeBlockTabsTrigger value="Mac/Linux">
          Mac/Linux
        </CodeBlockTabsTrigger>

        <CodeBlockTabsTrigger value="Windows">
          Windows
        </CodeBlockTabsTrigger>
      </CodeBlockTabsList>

      <CodeBlockTab value="Mac/Linux">
        ```bash
        export MEM0_API_KEY="your_mem0_api_key_here"
        export OPENAI_API_KEY="your_openai_api_key_here"
        ```
      </CodeBlockTab>

      <CodeBlockTab value="Windows">
        ```bash
        $Env:MEM0_API_KEY="your_mem0_api_key_here"
        $Env:OPENAI_API_KEY="your_openai_api_key_here"
        ```
      </CodeBlockTab>
    </CodeBlockTabs>
  </Step>

  <Step title="Run the example">
    Save the code above as `mem0_integration.py`, then run:

    ```bash
    python mem0_integration.py
    ```
  </Step>
</Steps>

Full source: [cookbook/11\_memory/integrations/mem0\_integration.py](https://github.com/agno-agi/agno/blob/8f36eaf2d18e91afa7b327eec66a3cd3685dcb87/cookbook/11_memory/integrations/mem0_integration.py)