Settings

Customize LocalChat's behavior, appearance, and AI response style.

Accessing settings

Open Settings by:

  • Clicking Settings in the sidebar
  • Pressing Cmd + , (comma)

Settings are organized into six categories: Models, Response Style, Cache & History, Appearance, Data & Export, and License.

Models

Manage your downloaded AI models and download new ones.

Model manager settings

Each model shows:

  • Name and description - What the model is best for
  • Status - Downloaded or available to download
  • Size - Download size and RAM requirements
  • Vision badge - If the model can process images

Actions

  • Download - Get a new model (shown for models not yet downloaded)
  • Load - Switch to this model for your conversations

See Model management for more details on choosing and using models.

Response style

Control how the AI generates responses. These settings affect the quality, length, and creativity of answers.

Response style settings

Simple mode

Toggle on to hide advanced options. When off, you can fine-tune the following settings.

Creativity

Controls how predictable or creative responses are.

  • Precise - Stick closely to likely answers (good for factual questions)
  • Creative - More varied and imaginative responses (good for brainstorming)

Default: Precise

Response length

How long the AI's answers can be.

  • Short - Brief, concise answers
  • Long - Detailed explanations

Default: Long

Focus

How closely the AI sticks to the most likely responses.

  • Narrow - More focused, predictable output
  • Wide - More diverse, exploratory responses

Default: Wide

Avoid repetition

Prevents the AI from repeating itself in responses.

  • Low - May repeat phrases
  • Normal - Balanced
  • High - Strongly avoids repetition

Default: Normal

Context

How much conversation the AI remembers. Higher values use more memory.

Options: 2K, 4K, 8K, 16K, 32K tokens

Default: 32K

Note: Changing context size requires restarting the model.

Speed mode

Use your Mac's GPU for faster responses.

Options:

  • Maximum Speed - Full GPU acceleration
  • Balanced - Mix of speed and power efficiency
  • Power Saver - Reduce GPU usage to save battery

Default: Maximum Speed

Cache & history

Manage conversation memory and storage for better performance.

Cache and history settings

Context cache

Shows whether conversation caching is active. Caching speeds up continued conversations by storing context.

  • Clear Cache - Remove cached data to free memory

KV cache

Speed up continued conversations by caching context between messages.

Default: On

Max cache size

How much memory to allocate for caching.

Options: 512MB, 1GB, 2GB, 4GB, Unlimited

Default: Unlimited

History limit

Number of messages sent to the model for context. More messages mean more context but use more memory.

Options: Last 5, Last 10, Last 20, Last 50, All

Default: Last 20

Appearance

Customize the look of LocalChat.

Appearance settings

Accent color

The main color used throughout the app for buttons, highlights, and UI elements.

Options: Default (dark teal), Red, Orange, Amber, Green, Teal, Blue, Indigo, Violet, Pink, Rose

Default: Default

User message color

Override the accent color specifically for your chat messages.

Options: Same as accent color options

Default: Default (uses accent color)

Data & export

Manage your conversation data and privacy.

Data and export settings

Export data

Download all your conversations as a JSON file for backup or analysis.

  1. Click Export All Conversations
  2. Choose where to save the file
  3. The JSON file contains all your chat history

Clear all chats

Permanently delete all conversations from LocalChat.

Warning: This action cannot be undone. All conversation history will be permanently deleted from your device.

  1. Click Delete All Conversations
  2. Confirm the deletion when prompted

License

Manage your LocalChat license.

  • View your current license status
  • Activate a new license key
  • Check license expiration (if applicable)