> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/janhq/jan/llms.txt
> Use this file to discover all available pages before exploring further.

# Settings & Configuration

> Configure Jan desktop app to match your preferences and optimize performance

## Accessing Settings

Open Settings in any of these ways:

* Click the **Settings** icon in the sidebar
* Press **Cmd/Ctrl + ,** (keyboard shortcut)
* Click your profile in the bottom-left corner

## General Settings

### Appearance

#### Theme

Choose your preferred color scheme:

* **Light** - Light background and dark text
* **Dark** - Dark background and light text
* **System** - Match your operating system theme

#### Font Size

Adjust text size for better readability:

* **Small** - Compact interface
* **Medium** - Default size
* **Large** - Easier reading

#### Language

Select your interface language:

* English
* Additional languages available

Changes apply immediately.

### Chat Preferences

#### Spell Check in Chat Input

Enable/disable spell checking while typing messages.

* **On** - Underline misspelled words
* **Off** - No spell checking

<Info>
  Spell check is provided by your operating system's built-in checker.
</Info>

#### Token Counter Display

Choose how to display token usage:

* **Compact** - Small indicator showing only essential info
* **Detailed** - Full breakdown of tokens and speed

Shows:

* Current context usage
* Generation speed (tokens/second)
* Total tokens vs. model capacity

### RAG (Retrieval Augmented Generation)

#### Enable Attachments

Turn document attachment features on or off.

* **Enabled** - Can attach and process documents
* **Disabled** - Document features hidden

#### Parse Mode

How Jan processes attached documents:

* **Prompt** - Ask each time how to process documents
* **Auto** - Automatically choose based on size and context
* **Inline** - Always insert content directly into prompts
* **Embeddings** - Always use vector database for semantic search

#### Max File Size

Maximum document size for attachments (in MB).

* Default: 10 MB
* Range: 1-100 MB

#### Auto Inline Context Ratio

When using Auto parse mode, documents smaller than this ratio of the model's context length will be inlined.

* Default: 0.75 (75%)
* Range: 0.1-1.0

## Models

Manage your downloaded and available AI models.

### Downloaded Models

View all models currently on your system:

* Model name and version
* Size and quantization
* Capabilities (text, vision, tools)
* Actions: Edit settings, Delete

### Model Hub

Browse and download models:

* Filter by capability, size, provider
* View model details and requirements
* Download with progress tracking
* Import models from Hugging Face or local files

<Card title="Learn More About Models" icon="brain" href="/desktop/models">
  Complete guide to downloading and configuring models.
</Card>

## Assistants

Create and manage custom AI assistants.

### Your Assistants

List of all your custom assistants:

* Name and avatar
* Description
* Edit or delete assistants

### Creating Assistants

Click **Add Assistant** to create a new one:

1. Set name and avatar
2. Write instructions
3. Configure parameters
4. Save

<Card title="Assistant Guide" icon="robot" href="/desktop/assistants">
  Learn how to create effective assistants.
</Card>

## Providers

Manage AI model providers and endpoints.

### LlamaCpp (Built-in)

Jan's default local inference engine:

* **Status** - Active/Inactive
* **Server Address** - Local endpoint
* **Port** - Network port (default: varies)

### External Providers

Add API-based providers:

1. Click **Add Provider**
2. Enter provider details:
   * Name
   * API base URL
   * API key (if required)
   * Model mappings
3. Test connection
4. Save

Supported provider types:

* OpenAI-compatible APIs
* Custom endpoints

### Provider Settings

For each provider, configure:

* **Timeout** - Request timeout duration
* **Proxy** - Proxy server settings (if needed)
* **TLS** - Certificate validation options

## Advanced Settings

### Jan Data Folder

Location where Jan stores all its data:

* Downloaded models
* Conversation history
* Attachments and embeddings
* Configuration files
* MCP servers

#### Viewing Location

Click **Open Folder** to view in file explorer.

#### Changing Location

1. Click **Change**
2. Select new folder location
3. Choose to move or copy existing data
4. Restart Jan

<Warning>
  Ensure the new location has sufficient storage space. Moving large models may take time.
</Warning>

<Card title="Data Folder Structure" icon="folder-tree" href="/desktop/data-folder">
  Learn about Jan's file organization.
</Card>

### Factory Reset

Reset Jan to default state:

1. Go to **Settings > Advanced**
2. Click **Factory Reset**
3. Choose what to reset:
   * Delete all threads
   * Delete all models
   * Reset all settings
4. Confirm reset

<Warning>
  Factory reset is permanent and cannot be undone. Export any important conversations first.
</Warning>

### Server Settings

#### Host

Network interface for local server:

* **127.0.0.1** - Localhost only (default, most secure)
* **0.0.0.0** - All network interfaces (for remote access)

<Warning>
  Exposing Jan on all interfaces allows network access. Only use on trusted networks with proper firewall rules.
</Warning>

#### Port

Network port for the local API server.

* Default: Auto-assigned
* Custom: 1024-65535

#### CORS Settings

Configure Cross-Origin Resource Sharing:

* **Allowed Origins** - Domains that can access the API
* **Trusted Hosts** - Whitelist for remote connections

## Remote Access Settings

Configure remote access to Jan:

### Tailscale Integration

Connect Jan to your Tailscale network:

1. Go to **Settings > Remote Access**
2. Click **Setup Tailscale**
3. Authenticate with Tailscale
4. Enable remote access
5. Share your Tailscale URL

<Info>
  Tailscale provides secure, encrypted access to Jan from anywhere without port forwarding.
</Info>

### Telegram Integration

Chat with Jan through Telegram:

1. Go to **Settings > Remote Access**
2. Click **Setup Telegram Bot**
3. Enter your Telegram Bot token
4. Select which model to use
5. Start chatting in Telegram

### WhatsApp Integration

Chat with Jan through WhatsApp:

1. Go to **Settings > Remote Access**
2. Click **Setup WhatsApp**
3. Scan QR code with WhatsApp
4. Configure settings
5. Start chatting

## MCP Servers

Manage Model Context Protocol servers.

### What are MCP Servers?

MCP servers provide external tools and data sources:

* File system access
* Database connections
* API integrations
* Custom functions

### Adding MCP Servers

1. Go to **Settings > MCP Servers**
2. Click **Add Server**
3. Choose configuration method:
   * **From Registry** - Pre-configured servers
   * **Manual** - Custom server configuration
4. Configure server settings:
   * Name
   * Command or URL
   * Arguments
   * Environment variables
5. Save and enable

<CodeGroup>
  ```json Example: Filesystem MCP theme={null}
  {
    "command": "npx",
    "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/allowed/dir"],
    "env": {}
  }
  ```

  ```json Example: GitHub MCP theme={null}
  {
    "command": "npx",
    "args": ["-y", "@modelcontextprotocol/server-github"],
    "env": {
      "GITHUB_PERSONAL_ACCESS_TOKEN": "your_token_here"
    }
  }
  ```
</CodeGroup>

### Managing Servers

For each MCP server:

* **Enable/Disable** - Toggle server availability
* **Edit** - Modify configuration
* **Delete** - Remove server
* **View Logs** - Check server output and errors

## Keyboard Shortcuts

View and learn all keyboard shortcuts:

### Application

* New chat: **Cmd/Ctrl + N**
* New project: **Cmd/Ctrl + Shift + N**
* Toggle sidebar: **Cmd/Ctrl + B**
* Zoom in: **Cmd/Ctrl + =**
* Zoom out: **Cmd/Ctrl + -**

### Chat

* Send message: **Enter**
* New line: **Shift + Enter**

### Navigation

* Search: **Cmd/Ctrl + K**
* Settings: **Cmd/Ctrl + ,**

<Card title="View All Shortcuts" icon="keyboard" href="/desktop/settings">
  Go to Settings > Shortcuts for the complete list.
</Card>

## Updates

### Auto-Update

Jan automatically checks for updates:

* **Notification** - Alert when new version available
* **Download** - Updates download in background
* **Install** - Prompt to install and restart

### Manual Update Check

1. Go to **Settings > About**
2. Click **Check for Updates**
3. Follow prompts to update

### Update Settings

* **Auto-download** - Automatically download updates
* **Notify only** - Show notification but don't auto-download
* **Disable** - Never check for updates

## Privacy & Security

### Data Storage

All data is stored locally:

* Conversations never leave your device
* Models run entirely on your computer
* No telemetry or tracking

### Optional Telemetry

Jan may collect anonymous usage data if enabled:

* App crashes and errors
* Feature usage statistics
* Performance metrics

No personal data or conversation content is collected.

**Toggle:** Settings > Advanced > Analytics

## Exporting Data

### Export Conversations

1. Open a thread
2. Click the menu (three dots)
3. Select **Export**
4. Choose format:
   * JSON - Machine-readable
   * Markdown - Human-readable
   * TXT - Plain text

### Backup Jan Data

Manually backup your Jan folder:

1. Go to **Settings > Advanced**
2. Click **Open Jan Data Folder**
3. Copy the entire folder to your backup location

<Tip>
  Regularly backup your Jan data folder, especially before major updates or system changes.
</Tip>