This guide covers every setting in the chatbot application. Click the ⚙️ gear icon to open the Settings panel, then explore each section below to understand what each option does.
📑 Table of Contents
- 🧠 Persona System Prompt
- 🧠 Memory Settings
- 🧩 Model Settings
- 🌐 Language Settings
- 🔊 Voice Settings
- 🎧 Audio Settings
- 🖥️ Display Settings
- 🖼️ Background Image
- ⚙️ Preferences
- 🐞 Debug Settings
- ❓ Help & Tutorial
🧠 Persona System Prompt
Control who your AI companion is—their identity, personality, and how they interact with you.
🆔 Core Identity Text Area
Define the AI's fundamental character. This is the unchangeable core of who they are—their name, role, and basic traits. Example: "You are Haru, a friendly and curious AI companion."
🎭 Custom Personality / Roleplay Text Area
Add custom personality traits, roleplay scenarios, or specific behaviors on top of the core identity. Example: "You are a tsundere childhood friend who secretly has a crush on the user."
💾 Save Persona Settings Button
Saves your persona configuration. Changes won't take effect until you save.
🔄 Reset Persona Button
Resets persona to the default settings, clearing any custom personality you've created.
💡 Tip: Keep your Core Identity concise (1-3 sentences). Use Custom Personality for detailed traits and scenarios.
🧠 Memory Settings
Control how much conversation history the AI remembers when generating responses.
🧵 Conversation Memory Size Slider (5-50)
The number of previous messages the AI includes for context. Higher values = better memory but more API usage. Default: 10 messages.
📊 Recommended Values:
| Range | Best For |
|---|---|
| 5-10 | Quick, lightweight conversations |
| 15-25 | Good balance for roleplay and ongoing stories |
| 30-50 | Maximum context for complex narratives (may be slower) |
🧩 Model Settings
Choose and manage your Live2D avatar models.
🎭 Select Model Dropdown
Choose from available Live2D models. Click the dropdown to see thumbnails and model names.
➕ Add Custom Model Input Fields + Button
Add your own Live2D model by providing:
- Model Name: Optional friendly name (auto-detected from URL if blank)
- Model URL: Direct URL to the
.model3.jsonfile (required) - Thumbnail URL: Optional image for the model selector
📂 Custom Models Dropdown
Select and manage custom models you've added. Shows model info and removal options.
♻️ Reset Model Position & Zoom Button
Resets the model's position and zoom level to defaults. Useful if you've dragged the model off-screen.
🗑️ Clear All Models Button
Removes all models currently displayed on screen.
📖 Open Model Gallery Button
Opens a fullscreen gallery showing all available models with thumbnails for easy selection.
🗑️ Clear Custom Models Button
Removes all custom models you've added (built-in models remain).
🌐 Language Settings
Configure the language for AI responses and interface.
🌍 Language Dropdown
Select from 193 supported languages. The AI will respond in your chosen language.
🔤 Show Transliteration (for JA, KO) Checkbox
When enabled, shows romanized pronunciation (romaji/romanization) under Japanese and Korean text.
🈯 Translate User Interface (AI) Checkbox
Translates all UI labels and buttons to your selected language using AI. May take a moment to load.
🔄 Reset Languages Button
Resets language settings to English (default).
🔊 Voice Settings
Control the AI's text-to-speech voice output.
🔈 Enable Voice (TTS) Checkbox
Master switch for voice output. When enabled, the AI speaks its responses aloud.
🎙️ Voice Dropdown
Select which voice the AI uses. Options include female voices in various languages. Set to "None" to disable TTS while keeping other voice controls visible.
📏 TTS Chunk Character Limit Slider (50-1000)
Maximum characters per TTS request. Lower values = more frequent but shorter audio clips. Default: 300.
▶️ Play / ⏸️ Pause / ⏹️ Stop Buttons
Manual controls for voice playback. Play resumes, Pause halts temporarily, Stop cancels all queued speech.
💡 Tip: The voice automatically matches your language when you switch languages. Languages without native voices use the closest phonetic alternative.
🎧 Audio Settings
Ambient audio and background music.
📻 Radio Stream Play/Pause + Volume Slider
Toggle an anime music radio stream (Listen.moe). Adjust volume with the slider (default: 25%).
🖥️ Display Settings
Customize the visual appearance of the interface.
⏰ Show Clock Checkbox
Displays a digital clock on the screen.
🪟 Chatbox Opacity Slider (0.1-1.0)
Controls transparency of the entire chat container. Lower = more transparent. Default: 0.9.
💬 Message Bubble Opacity Slider (0.1-1.0)
Controls transparency of individual message bubbles. Default: 0.3 (quite transparent for aesthetic).
🖼️ Background Image Opacity Slider (0-1.0)
Controls how visible the background image is. 0 = hidden, 1 = fully visible. Default: 1.0.
🖼️ Background Image
Customize the scene behind your AI companion.
✨ AI Background Prompt Text Input + Button
Describe a scene and click "Generate Background" to create an AI-generated image. Example: "tranquil starry night sky, minimal, soft light"
🌐 Generate from Conversation Button
Automatically generates a background based on the current conversation context.
🔗 Custom Background URL Text Input + Button
Set a background using any image URL. Click "Set Background" to apply.
🧹 Clear Background Button
Removes the current background, returning to the default dark theme.
📚 Background Library Gallery
View and manage all generated or added backgrounds. Click to apply, use Select mode to delete multiple.
📐 Background Fit Button Group
Control how the background image is displayed:
- Contain: Full image visible, may have letterboxing
- Cover: Fills screen, may crop edges
- Stretch: Stretches to fill (may distort)
- Top/Center/Bottom: Aligns the image vertically
- Fit Width/Height: Scales to fit one dimension
⚙️ Preferences
General application behavior settings.
🪄 Always show Settings on page load Checkbox
When enabled, the Settings panel opens automatically every time you load the page. Otherwise, it remembers whether you left it open or closed.
⏱️ Include current time in context Checkbox
Provides the AI with the current date and time so it can give time-aware responses like "Good morning!" or reference the current day.
🔋 Include battery status in context Checkbox
Lets the AI know your device's battery level and charging state. The AI might say "I see you're running low on battery!"
🧬 Load multiple models at once Checkbox
When enabled, selecting a model adds it to the screen without removing existing ones. Great for having multiple characters interact!
🐞 Debug Settings
Tools for developers and troubleshooting.
🧰 Enable Debug Panel Checkbox
Shows a draggable debug panel with real-time logs of system events, AI responses, and errors.
📜 Auto-scroll Log Checkbox
Automatically scrolls the debug log to show the newest entries.
🧹 Clear Debug Log Button
Clears all entries in the debug panel.
💬 Show Live2D Debug Logs Checkbox
Shows detailed logs about Live2D model loading, animations, and interactions.
🧠 Show AI Debug Logs Checkbox
Shows logs about AI requests, responses, and processing.
🔊 Show TTS Debug Logs Checkbox
Shows logs about text-to-speech processing, chunking, and playback.
🔌 Test Fallback AI Checkbox
Forces the app to use the offline fallback AI instead of the main API. Useful for testing offline behavior.
⚠️ Note: Debug settings are primarily for developers. Enabling all logs may affect performance.
❓ Help & Tutorial
📘 Launch Tutorial Button
Opens an interactive walkthrough that explains the main features of the application.
📋 Quick Reference Table
| Setting | Type | Default | What It Does |
|---|---|---|---|
| Memory Size | Slider | 10 | Messages AI remembers |
| Language | Dropdown | English | AI response language |
| Voice | Dropdown | Auto | TTS voice selection |
| TTS Chunk Limit | Slider | 300 | Max chars per speech chunk |
| Chatbox Opacity | Slider | 0.9 | Chat container transparency |
| Message Opacity | Slider | 0.3 | Message bubble transparency |
| Background Opacity | Slider | 1.0 | Background image visibility |
| Radio Volume | Slider | 0.25 | Background music volume |
🎉 Pro Tip: All settings are automatically saved to your browser's local storage. They'll persist across sessions until you clear your browser data.