Aller au contenu principal

Chat Preferences

Chat Preferences gives you control over how messages are sent, how the interface is laid out, and how content is rendered. Changes here affect every conversation.


Opening Chat Preferences

Click your user name or avatar at the bottom of the left sidebar → SettingsChat tab.


Input Behavior

Enter to Send

When on: pressing Enter sends your message. Use Shift+Enter to insert a line break.

When off: Enter inserts a line break. Use the send button to submit.

Turn this off if you frequently write multi-line messages and find accidental sends disruptive.


Layout Controls

SettingWhat It Does
Maximize Chat SpaceExpands the message area to use more of the screen width
Center Chat Input on LandingCenters the input field when no conversation is open — the default landing view
Show Scroll ButtonShows a button to jump to the bottom of the conversation when you are scrolled up
Font SizeAdjusts the size of message text (Small / Medium / Large or equivalent options)
Chat DirectionSets the text direction: LTR (left-to-right) or RTL (right-to-left)

Content Rendering

SettingWhat It Does
Show CodeControls whether code blocks are rendered with syntax highlighting
LaTeX ParsingWhen on, mathematical expressions wrapped in $$...$$ are rendered as formatted equations
Show ThinkingWhen on, extended reasoning steps are displayed for models that support it; when off, only the final response is shown
Modular ChatEnables an alternative chat layout with modular message presentation

Conversation Behavior

SettingWhat It Does
Save DraftsPreserves unsent message text when you switch to another conversation; the draft is restored when you return
Save Badges StateKeeps badge visibility consistent across sessions (affects notification indicators on conversation list items)
Default to Temporary ChatWhen on, new conversations are temporary by default — they are not saved to your conversation history
Fork Conversation SettingsControls how conversation branching behaves when you edit a message or start from a previous point

Note on "Show Thinking"

The Show Thinking toggle only has a visible effect when using a model that supports extended reasoning. On standard models, the toggle has no impact on what is displayed.