How to install and access the Mem0 Dashboard in a local deployment?

[vvolt-ai]

We’re setting up a local instance of Mem0 to evaluate it for a production environment. So far, the core service is running fine, and we’re able to use Qdrant directly to inspect and manually edit stored memories. However, we haven’t been able to find any clear documentation or examples on how to install and access the Dashboard UI.

What we tried

Installed Mem0 locally following the docs (Docker + manual approaches).

Explored the repo for any dashboard-related subproject, Dockerfile, or environment variable (no luck).

Looked for hints in the GitHub Issues / Discussions and codebase for dashboard, ui, frontend, or vite (found some scattered references but nothing actionable).

Checked Qdrant and confirmed memories are stored correctly — so the backend seems operational.

Where we’re stuck
We can’t find a supported way to launch or connect to the dashboard, either as part of the main container or a separate service. It’s unclear whether the dashboard is:
A separate repo / package we need to clone,
An internal tool not currently released,
Or something that should start automatically but isn’t due to misconfiguration.

Questions

What is the correct way to install and run the Mem0 dashboard on a local setup?
Is there any separate repo or branch we should be pulling for the dashboard UI?
Are there specific environment variables or ports we should expose to make it work?
If the dashboard isn’t publicly available yet, is there a recommended alternative for visualizing and managing memories?

Environment
OS: Windows 11
Deployment: Local (Docker + manual)
Vector DB: Qdrant (running fine)
Mem0 version: latest as of Oct 2025

Why this matters
The dashboard seems like a key piece to interact with and visualize stored memories beyond raw vector inspection. Having it working locally would make it much easier to evaluate Mem0 and potentially contribute back.

Any guidance or pointers would be greatly appreciated!

[parshvadaftari]

Hey @vvolt-ai
Thanks for exploring Mem0 and setting it up locally!

At the moment, there isn’t a dashboard UI available for Mem0. All interactions and memory management are handled through the API or directly via the connected vector database (like Qdrant).

We understand how a visualization layer could be useful, and it’s something the community has expressed interest in — so we’ll consider adding clearer guidance or future tooling around that.

Appreciate you taking the time to evaluate Mem0 and share your feedback!

[dabing1022]

Can I understand that the UI part in the OpenMemory folder is the dashboard?

[parshvadaftari]

No the ui part in the OpenMemory folder is for the OpenMemory dashboard and not for the mem0 dashboard.

[mecampbellsoup]

+1, this would be a nice feature OOTB. A Django admin-esque way to "see" memories - what mem0 does out of the box, how its view of memories changes over time, etc.

[xXMrNidaXx]

The Mem0 Dashboard is currently a cloud-only feature, not available for self-hosted deployments.

Current status:

• Dashboard UI is part of mem0.ai cloud platform
• No open-source dashboard in the OSS repo
• This is a common ask from self-hosters

Alternatives for local visualization:

1. Qdrant Dashboard (built-in)

# Qdrant UI at http://localhost:6333/dashboard
docker run -p 6333:6333 qdrant/qdrant

View collections, search vectors, inspect payloads.

2. Build a simple dashboard

import streamlit as st
from mem0 import Memory

memory = Memory(...)

user_id = st.text_input("User ID")
if user_id:
    memories = memory.get_all(user_id=user_id)
    for m in memories:
        st.write(f"{m.created_at}: {m.text}")
        if st.button(f"Delete {m.id}"):
            memory.delete(m.id)

3. Langfuse for observability

from langfuse import Langfuse

# Track memory operations
langfuse.trace(name="memory_add", input=messages, output=result)

4. Direct Qdrant API

from qdrant_client import QdrantClient

client = QdrantClient("localhost", port=6333)
memories = client.scroll(collection_name="mem0", limit=100)

Feature request:
Consider opening an issue for open-source dashboard — community would likely contribute.

We build memory visualization tools at Revolution AI — Streamlit + Qdrant dashboard is quick to set up.