Text-to-Speech (TTS)

Uses the browser's built-in Web Speech API to read AI responses aloud.

How to Use

Click the 🔊 read-aloud button to start playback; click the same button again to stop.

Text Preprocessing

Before playback, Momo cleans the response text:

Removes thinking blocks (<think>, <thinking>, <thought>)
Strips HTML tags and Markdown syntax
Removes emoji and decorative symbols
Applies Traditional / Simplified Chinese conversion based on your language preference
Converts semantic symbols (such as &, %, +, =, @, common currency symbols) into natural words before speech — e.g. Tom & Jerry → Tom and Jerry
Other punctuation and decorative symbols are not sent to the speech engine; they are used only for segmentation and pauses (Reading rhythm)

Settings (Settings → Text-to-Speech)

Voice Selection

Choose a voice from the list of available system and browser voices.

💡 More Voices You can download and install additional voices in macOS or Windows system settings. Once installed, they usually appear in this list.

Rate

Adjust the speech speed.

Pitch

Adjust the voice pitch.

Preview

Use the preview button to listen to the current settings before applying them. Preview uses the same preprocessing logic as sidebar playback.

Reading Rhythm

Text is split by punctuation with short pauses between segments so longer replies sound more natural. Punctuation itself is not spoken — it only affects pause timing.

Auto-read AI Replies

Settings include an Auto-read AI replies toggle (off by default). When enabled, Momo reads each completed AI reply automatically. You can still use the sidebar 🔊 button to play or stop manually.

Auto-read and manual playback share the same preprocessing: semantic symbols become natural words, while other punctuation is used only for pauses and is not spoken aloud.