Serenade Changelog

See our blog post for more.
Dedicated dictation mode, which defaults to inserting words that are heard. You can enable with start dictating and disable with stop dictating.
Natural English model for formatting prose text.
Language switcher to control which model is used to format your code and text.
Integration with macOS and Windows accessibility APIs, to edit text without a plugin.
Dictation box to edit text when both plugins and accessibility APIs are unsupported.
New model for voice activity detection, so Serenade works better with background noise.
New model for better understanding command intent and resolving ambiguities.
Switch to a WebSocket-based API for all network communication, for compatibility.

Support for creating plugins via the Serenade Protocol
Switch to the open-sourced serenade-driver for system integrations.
Support for third-party npm packages in custom commands.
Microphone settings page now shows current microphone volume, so you can more easily test your microphone.
Added API function evaluateInPlugin to call VSCode APIs from custom commands, rather than relying on keyboard shortcuts.
Added API functions for click and setCursorPosition across macOS, Windows, and Linux.
Added API functions for mouseDown and mouseUp.
Added API function focusOrLaunch for focusing an app if it's running, and launching it if not.
Support for non-English keyboard layouts across macOS, Windows, and Linux.
Microphone fixes to improve audio quality and accuracy.
Added complete symbol, key, and selector lists to the documentation.

New mode indicators to make it easier to see which Serenade commands are supported, and which are not.
Fix issues causing Serenade to lag on long commands.
Support for unicode characters in source code.
Save microphone settings between app launches.
Add entry for using the default system microphone.
Fix issue where "repeat" didn't work after selecting a different alternative.
Fix issue where the Serenade app didn't always choose the nearest Cloud server.
serenade.include API method for including a custom commands file from another.
Add symlink support for ~/.serenade/scripts.

Support for Mini-mode.
Support for connecting through proxy servers.
Support for customizing which styler is used for the style file command. Stylers include Black (Python), yapf (Python), google-java-format (Java), clang (C++), and more.
Support for prepositional phrases after commands, like insert foo after bar, as an alternative to chaining commands like end of bar insert foo. (Both work!)
More reliable audio streaming for both Serenade Cloud and Serenade Pro.
More robust detection for when a command has finished being spoken.
Fixes for focus across all platforms.
Fixes for surround with, and introduction of commands like surround with div tag
Fix change commands that weren't working when changing multiple words.
Fixes for selecting or changing text that also matches the name of a Serenade command (e.g., "next word paste" will be parsed as "go to next word, then paste" rather than "go to the next word called paste").
Commands like go to and change match any case, so you don't have to explicitly specify one.
Add a border to the Windows and Linux applications to make the app stand out more on white backgrounds.

General availability of Serenade Pro.
New tutorials for Python, Java, and JavaScript, covering frameworks like Flask, React, and Express.
Added support for C and C++.
Apple Silicon compatibility.
When adding code at a blank line, an extra newline won't be created.
Fix chained commands in Visual Studio Code.
Support for custom word pronunciations.
Support for clicking native macOS buttons via click commands.
next line creates a newline if the cursor is at the end of a file.
Updated our speech engine to improve accuracy across all languages.
Updated our speech-to-code AI platform to enable more ambiguous descriptions of code.

Added the change all command for changing all instances of a string in a file.
Added the shift command for moving objects around (e.g., shift argument right and shift line up).
Added selectors for before and after, so you can say commands like after foo type bar.
Added the forward delete key for triggering a forward delete on macOS.
Improvements to background noise handling.
Fixed issue with chaining commands in VS Code (e.g., next tab line one).

Added first-class support for JetBrains IDEs via Serenade for JetBrains
Added first-class support for the Kotlin programming language
Added first-class support for the Dart programming language
Added VS Code debugger support with commands like start debugging, step over, and toggle breakpoint
Added the duplicate command for copying and pasting selectors
Added the surround command for surrounding selectors with text
Added the rename command for changing the name of selectors
Added the join lines command for combining multiple lines into one
Added the <style> <selector> command for toggling the text style of selectors (e.g., capitalize next two words)
Added the switch window command for toggling among multiple windows of the same application in macOS
Added the title case text style for formatting text In Title Case
Added shorter camel and pascal aliases for camel case and pascal case
Added stop listening as an alias for pause
Fixed issue causing the microphone to cut off commands too early

Visual refresh of client application and website
Added dictate command for speaking text without programming symbols
Added system command for typing text into native OS fields rather than the editor field
Fixed open command for VS Code remote projects
Fixed next/previous behavior in tutorial mode
Fixed parsing of files with comments inside enclosures