Settings ​
This page documents every option available on the Settings page of Page Assist. To open it, click the extension icon, open the Web UI, and go to Settings → General Settings (the gear icon).
The Settings page is organized into the sections described below.
Web UI Settings ​
These options control the general behaviour of the chat experience.
| Setting | Description |
|---|---|
| Language | Changes the display language of Page Assist. |
| Resume the last chat when opening the SidePanel (Copilot) | Reopens your most recent SidePanel conversation instead of starting a new one. |
| Enable Chat with Website by default (Copilot) | Starts the SidePanel in "Chat with Website" mode automatically. |
| Resume the last chat when opening the Web UI | Reopens your most recent Web UI conversation on launch. |
| Hide the current Chat Model Settings | Hides the quick model-settings panel inside the chat view. |
| Restore last used model for previous chats | Restores the model that was used when you reopen an old chat. |
| Send Notification After Finishing Processing the Knowledge Base | Shows a browser notification when knowledge-base indexing completes. |
| Generate Title using AI | Automatically generates a chat title from the conversation. |
| Enable or disable Ollama connection status check | Toggles the periodic Ollama health check. |
| Enable wide screen mode | Uses a wider chat layout, useful for large screens. |
| Open Reasoning Collapse by default | Expands the reasoning/thinking block automatically for new messages. |
| Show Thinking Mode State in Forms | Displays the thinking-mode state inside model setting forms. |
| Use Chat Bubble for User Messages | Renders user messages inside chat bubbles. |
| Automatically Copy Response to Clipboard | Copies each AI response to the clipboard as soon as it finishes. |
| Enable Markdown formatting for User messages | Renders your own messages as Markdown. |
| Copy as Formatted Text | Copies responses as rich/formatted text rather than plain text. |
| Enable Tab Mentions (@tab) 🧪 | Lets you reference open browser tabs with @tab mentions. |
| Paste Large Text as File | Converts large pasted text into an attached file instead of inline text. |
| Default OCR Language | The default language used when extracting text from images. |
| Enable Temporary Chat in SidePanel by default | Starts SidePanel chats as temporary (not saved) by default. |
| Enable Temporary Chat in Web UI by default | Starts Web UI chats as temporary (not saved) by default. |
| Remove Reasoning Tag from Copied Text | Strips the reasoning/thinking block when copying a response. |
| Show the 'Summarize' button on YouTube videos 🧪 | Adds a summarize button to YouTube video pages (not available on Firefox). |
| Hide Reasoning Widget from AI Messages | Hides the reasoning widget from AI responses. |
| Persist Chat Input (Save unsent messages) | Keeps unsent text in the input box across sessions. |
| Enable Message Queue While Streaming | Lets you queue new messages while a response is still streaming. |
| Optimize Chat UI for Small Screens | Adjusts the chat layout for smaller screens. |
| Enable Text Wrapping in Markdown Tables | Wraps long cell text in Markdown tables instead of scrolling. |
| Show 'Show more' for large human messages | Collapses long user messages behind a "Show more" toggle. |
| Sidebar Position | Places the sidebar on the Left or Right. |
| Show MCP Servers Toggle in Chat | Shows the MCP servers toggle in the chat input area. |
| Use Agent Web Search (model decides when to search and fetch) | Lets the model decide when to search the web and fetch pages. |
| Require approval before running MCP tools | Prompts you to approve each MCP tool call before it runs. |
| Show Provider Name in Model List | Displays the provider name next to each model in the model picker. |
| Default Prompt for SidePanel (Copilot) | Default system prompt applied to new SidePanel chats. |
| Default Prompt for Web UI | Default system prompt applied to new Web UI chats. |
| Change Theme | Switches between Light and Dark mode. |
Manage Web Search ​
Configure the internet search behaviour. The available fields change depending on the selected search engine.
| Setting | Description |
|---|---|
| Search Engine | The search provider to use (Google, DuckDuckGo, SearXNG, Brave API, Tavily API, Exa, Firecrawl, Ollama Search, Kagi, Perplexity, and more). |
| SearXNG URL | The base URL of your SearXNG instance (shown when SearXNG is selected). |
| Google Domain | The Google domain to query (shown when Google is selected). |
| Brave / Tavily / Exa / Firecrawl / Ollama Search / Kagi / Perplexity API Key | The API key for the selected API-based provider. |
| Perform Simple Internet Search | Uses a faster, lightweight search that skips visiting individual pages. |
| Total Search Results | How many search results to retrieve. |
| Visit the website mentioned in the message | Fetches and reads URLs that appear directly in your message. |
| Internet Search ON by default | Turns on internet search for every new chat automatically. |
| Domain Filter List | Only show results from the listed domains. |
| Blocked Domains | Exclude results from the listed domains. |
For a full walkthrough of internet search, see Internet Search.
Speech-to-Text Settings ​
| Setting | Description |
|---|---|
| Speech Recognition Language | The language used for voice input/transcription. |
| Auto Submit Voice Message | Automatically sends the message when you stop speaking. |
| Auto Stop Timeout (ms) | How long to wait (in milliseconds) before automatically stopping recording. |
Text-to-Speech Settings ​
| Setting | Description |
|---|---|
| Enable Text-to-Speech | Turns on reading AI responses out loud. |
| Auto play voice response after completion | Plays the spoken response automatically when a message finishes. |
| Text-to-Speech Provider | The TTS engine to use: Browser TTS, ElevenLabs, OpenAI TTS, or Mistral TTS. |
| Text-to-Speech Voice | The voice used for playback (options depend on the selected provider). |
| API Key / Base URL / TTS Model | Connection details for ElevenLabs, OpenAI TTS, or Mistral TTS providers. |
| Response Splitting | How responses are split before being spoken: None, Punctuation, or Paragraph. |
| Enable SSML (Speech Synthesis Markup Language) | Interprets SSML tags in the response for finer playback control. |
| Remove Reasoning Tag from TTS | Skips the reasoning/thinking block when reading a response aloud. |
| Playback Speed | The speed at which speech is played back (e.g. 1 for normal). |
System Settings ​
| Setting | Description |
|---|---|
| Font Size 🧪 | Increases or decreases the interface font scale. |
| Set Default Action for Extension Icon Clicks 🧪 | Choose whether clicking the toolbar icon opens the Web UI or the SidePanel. |
| Set Default action for Context Menu 🧪 | Choose whether the right-click context menu opens the Web UI or the SidePanel. |
| Sync Custom Models, Prompts for Firefox Private Windows 🧪 | (Firefox only) Syncs custom models and prompts into private/incognito windows. |
| Enable Browser Storage Sync (Sync settings across devices) | Syncs your settings across devices using browser storage sync. |
| Show Web UI Button in Side Panel | Adds a button to open the Web UI from inside the SidePanel. |
| Chat Background Image 🧪 | Upload a custom background image for the chat view. |
| Export All Data | Export chat history, knowledge base, prompts, MCP servers, model configs, and settings to a JSON file. You can choose which sections to include. |
| Import All Data | Import data from a Page Assist export file (Open WebUI exports are also supported). You can choose which sections to import. |
| System Reset | Clears all data. This cannot be undone. |
TIP
Settings marked with 🧪 are experimental and may change or behave differently across browsers.