Greasy Fork

Refines AI responses and improves lesson auto-completion logic.

Added am audio player to the TTS playground.

Added TTS playground.

Fixed a bug on the print page. Refactored the code structure.

Added a progress text for the lesson audio generation. Enhanced the progress bar.

Added more random options for google cloud TTS.

Changed the google cloud voice list loading logic.

Added the Google Cloud as a new TTS Provider. It's stable and has more varied voices.

Implemented lesson audio generation and refactors core setup logic.

Enhanced summary prompts and fixed a bug related to the quick summary loading.

Fixed lesson summary to update when lesson is changed without refreshing.

Enhanced the prompts.

Added a regenerate response button to the last bot message.

Added a copy message button to bot messages.

Added a read the message button to bot messages. It utilizes the AI TTS.

The read the message button in a word-response only reads the word and the example sentence.

Modified the style of the buttons in bot messages.

Refined AI prompt logic, improved chat button styling.

Added a setting to prepend lesson summary.

Refined the lesson summary prompts.

Enhanced prepended summary style

Added a setting to prepend lesson summary.

Improved chat button styling.

Enhanced the prompts for more consistent responses.

Modified the style of the buttons in bot messages.

The read the message button in a word-response only reads the word and the example sentence.

Added a read the message button to bot messages. It utilizes the AI TTS.

Added a copy message button to bot messages.

Added a regenerate response button to the last bot message.

Enhanced the prompts.

Fixed lesson summary to update when lesson is changed without refreshing.

Fixed a bug related to getting the lesson text.

Made LLM have a lesson summary in its context.

Fixed a bug that cursor is moving to the chat input box after the first response is generated.

Made the prompts use system role and refined chat UI logic.

Added a donation button.

Implemented LLM response streaming to enhance the chat experience.

Highlighted LingQ words in print and improved vocabulary table.

Improved print layout by adding page break before vocabulary table.

Added a vocabulary table to the print page.

Ensured translation is shown when a lesson is changed without a refresh.
Corrected the approximate cost calculation for cached tokens

Added a new model (gemini 2.5 flash light)

Added an autoplay in sentence view.

Modified the showTranslation function's logic.

Edited CSS for the minimized video player.

Added the checkbox that shows translation automatically.

Added the focusPlayingSentence checkbox.

Fixed the problem that the playing sentence focusing function isn't working with the translation is turned on.

Added the translation text color picker to the stting page.

Changed the style of the send chat button.

Enhanced the prompts. Fixed the wrongly calculated input tokens of gemini chat.

Modified the style of chat message. Fixed a bug related to the widget that doesn't have .is-fixed.

Integrated the youtube player simplification script into the code (Automatically collapse More videos popup in the player).

Minor UI update.

Added the gpt-4.1 as an option to the model list. Enhanced the prompt. Applied retry to the google TTS. Changed the logic of the TTS button.

Fixed the problem that AI-TTS is not working when chat widget is turned off.Fixed the problem that AI-TTS is not working when chat widget is turned off.

Improved Google TTS Prompt. Added a random voice option.

Add Google TTS option.

Modified the prompts. LLM console logs now display the corresponding costs.

Changed the default color palette to the same as the original LingQ setting.

Made the playing sentence focusing scroll smooth.

Fixed the bug on the tts button.

Better logging. Fixed the bug on the tts button.

Fixed the bug on the tts button.

Better logging. Fixed the bug on the tts voice selector

Fixed the issues occurring when moving between the lessons without refreshing. The existing TTS will be replaced with AI TTS when clicked.

Optimized the waitForElement for performance reason. Improved the prompt for sentence.

Huge code refactoring for performance and clean code reasons.

Wrote logging codes for easier debugging, fixed prompts. Added the keyboard shortcut on/off option.

Applied formatting to the input word/sentence and context.

Added a 'Meaning Copied' popup message.

Meaning of the selected word/sentence is copyed to clipboard.

Improved formatting of bot messages in the chat widget by changing the prompt. Add a filter that filters TTS requests with more than 1000 characters of text.

Add customizable keyboard shortcuts.

Update CSS variables and styles for better video player in the audio layout.

Enhance DOM traversal for leaf nodes and safeguard audio context state. Fix bug related to the reference word changing.

Optimized DOM element checks using matches. Fixed the bug related to the reference word changing.

Fixed the issue where the characters (`.`, `,`, `?`, and `\n`) were omitted from the selected text.

Improved styles. Prevented the TTS audio could being played multiple times at the same time.

Added chat widget for all widgets. Added a loading message while a bot message is loading.

Added widgetWidth slider, Refined CSS variable naming.

Refine the chat widget styles. Make a keyboard shortcuts for the chat widget (esc). Refined the prompt for more aesthetic and structured output.

Removed the scroll bar of the chat widget container, and changed the size of that.

Decreased the volume of the TTS. Modified the selector of the word to select words also. Modified the chat widget prompt. Added more model options Disabled the thinking option for the gemini model.

Modified the prompt for the chat widget.

Updated the CSS for clearer UI elements.

Separated the logic of the widget and the TTS. Modified the prompt. Fixed the bug that the TTS and the chatbot widget are not reloaded when selecting a sentence while selecting another one.

Enhanced the prompts, fixed the bugs related to the tts.

Added an AI-based TTS option.

Rewrote the system prompt. Make gemini use system prompt.

Changed the default model. Made the "off mode" also affected by the style settings.
Refactored the ChatWidget code. Added the checkbox for enabling/disabling the chat widget.

Changed the default model. Made the "off mode" also affected by the style settings.

Provide a chatbot widget.

Modified the finish lesson condition.

Provide the ratio of both LingQ and KnownWords of a lesson.

Refactored the setupCourse function.

Refactored the setupSliderObserver function.

The lesson's progress is updated according to the progress of the video.

Disable video closing when pressing the ESC key while typing.

Fixed a bug in the toast pop-up.

Encapsulated the download function call with a handler.

Updated progress bar texts

You can download every known/unknown word/LingQs.

Added the download words popup. You can download every known word.

Fixed a bug in the video header selector.

Fixed the bugs related to the style/color mode selector event listeners.

Fixed the popup(lesson info, setting, etc.)-related issues.

Adjusted the scroll offset

Fixed scroll-related issue

Fixed z-index-related bug