---
title: Deploy LibreChat with AI on Blossom
url: "/deploy/librechat/"
canonical_url: https://docs.blossom88.com/deploy/librechat/
markdown_url: https://docs.blossom88.com/deploy/librechat.md
summary: Deploy LibreChat with AI on Blossom. Run a private chat workspace with your
  model provider, managed MongoDB, and saved conversations.
categories:
- deployment
- ai
---

# Deploy LibreChat with AI on Blossom

[LibreChat](https://www.librechat.ai/) is a chat interface for working with
AI model providers. Sign in, choose a configured model, and keep conversations
in your own deployment.

It uses Node.js and Express with a React interface, and MongoDB for accounts,
conversations, and settings. Browse the
[source on GitHub](https://github.com/LibreChat-AI/LibreChat).

This setup runs the chat application on regular App Capacity with managed
MongoDB. Your provider runs the model; LibreChat doesn't need a local model
or a GPU.

[Install the Blossom CLI and skill](/learn/getting-started/install/)
on your agent's machine and choose a Blossom Cloud Location when you log in.
You can also use
[Blossom BYOC (bring your own cloud)](/learn/getting-started/choose-your-cloud/#blossom-byoc).

## Ask Your Agent

> **Ask your AI agent:** Deploy LibreChat on Blossom with managed MongoDB. Keep it private and help me connect my model provider.

> Follow this guide: https://docs.blossom88.com/deploy/librechat/



Your agent checks the current
[deployment requirements](https://www.librechat.ai/docs/local/docker), selects
a compatible maintained release, and saves the configuration. Supply your model
provider's API key through a private environment file.

| Part | Purpose |
|---|---|
| Chat Service | Sign-in, model selection, and conversations |
| Managed MongoDB | Accounts, messages, and saved settings |
| Your model provider | Generates responses |

Keep the site behind **Blossom Account access**. Create the initial owner,
then disable open registration unless you intend to accept new accounts.
Preserve authentication and credential-encryption secrets through updates.

## Start a Conversation

> **Ask your AI agent:** Help me send a message, save my preferred chat settings, and reopen the conversation.

> Follow this guide: https://docs.blossom88.com/deploy/librechat/



Select a configured model and send a short message. Save a preset for settings
you want to reuse. Sign out and back in to confirm that the conversation and
preset are saved.

![LibreChat displaying a saved conversation with a configured model.](/assets/images/guides/librechat/conversation.png)

<details class="optional-reading" markdown="1">
<summary>Commands behind the deployment</summary>

Create a prebuilt-image Environment, create MongoDB, and link it to the app:

    blossom app:create --data app.yml
    blossom environment:create --data environment.yml
    blossom database_instance:create --data mongodb.yml --follow
    blossom environment_link:create --data mongo-link.yml
    blossom env_var:push --environment environment-100 --file .env.production
    blossom service:create --data chat-service.yml
    blossom environment:deploy --id environment-100 --follow

Use a MongoDB Environment Link named `MONGO_URL`. At startup, export it as
`MONGO_URI`, the name LibreChat expects. Keep the supplied connection options
so the driver uses Blossom's private TLS connection.

Set `DOMAIN_CLIENT` and `DOMAIN_SERVER` from `BLOSSOM_SERVICE_URL`, or your
custom HTTPS domain. Listen on port `3080` and use `/health` for the HTTP check.
Preserve the image's supported startup command and non-root user.

Save `librechat.yaml` with the deployment and provide it at startup. Enable the presets menu
with `interface.presets: true`. Configure
only your intended providers and models. Keep `CREDS_KEY`, `CREDS_IV`,
`JWT_SECRET`, and `JWT_REFRESH_SECRET` stable, with formats required by the
selected release. Keep provider keys in environment variables.

This chat-only setup keeps records in MongoDB, logs to standard output, and
disables uploads and scheduled chats. Document retrieval, local files,
conversation search, and scheduled work
need their own storage and service configuration before you enable them.

</details>

## Keep Conversations Through Updates

> **Ask your AI agent:** Update LibreChat and check that I can sign in, reopen my conversation, and send another message.

> Follow this guide: https://docs.blossom88.com/deploy/librechat/



Keep MongoDB, deployment configuration, and secrets. After Container
replacement, sign in again and confirm your previous messages and preset
remain, then continue the conversation.

This guide tests one chat instance. Follow LibreChat's current shared-state
requirements before adding replicas or background features.

## What You Get

- A private chat interface for your chosen model provider
- Accounts and conversation history in managed MongoDB
- Saved chat presets across Container replacement
- Regular App Capacity for the chat Service
