Serenade

Serenade is an open-source, speech-to-code voice assistant built specifically for developers. It lets users write, edit, and navigate code entirely through natural voice commands, and is available as a free download with its full source code published on GitHub under the Apache License 2.0. The project supports both cloud-based and fully local (on-device) processing, giving developers control over where their voice commands and source code are handled.

What It Is

Serenade is a developer-focused voice coding tool that translates natural speech into code and editor actions. Unlike general-purpose dictation software, its speech engine is designed from the ground up for programming workflows — understanding syntax, code structure, and developer commands rather than just transcribing words. The core platform, online services (speech engine, code engine), and model training code are all published in the Serenade monorepo on GitHub under the Apache-2.0 license.

Cloud and Local Processing Modes

One of Serenade's key architectural choices is its dual-mode processing. Users can run the speech-to-code engine in the cloud to minimize local resource usage, or run it entirely on-device so that voice commands and source code never leave the machine. The project page describes this as a privacy and performance tradeoff that is fully in the user's control.

Integrations and Plugin Ecosystem

Serenade integrates with a wide range of developer tools through dedicated plugins, each maintained in separate repositories:

• Editors/IDEs: VS Code, Atom, JetBrains (IntelliJ, PyCharm, WebStorm, Android Studio), Hyper, iTerm2
• Languages: Python, JavaScript, TypeScript, Java, C/C++, Dart, Bash, HTML, CSS, Sass, Markdown
• Platforms/Services: Chrome, Slack, Discord, Gmail, GitHub, GitLab, JIRA, Linear, Jupyter, Google Colab, Repl.it, LeetCode

The plugin architecture means Serenade slots into existing workflows rather than replacing them. It also provides an open protocol for building custom voice commands and community-shared plugins.

Custom Voice Commands

Serenade exposes a scripting API that lets developers write custom voice commands in JavaScript. The homepage shows an example where a user defines a "build" command that focuses the terminal, clears it, types yarn build, and presses return — all triggered by a single spoken phrase. Community-created customizations can also be shared and reused.

Open-Source Status and Repository Activity

The Serenade monorepo on GitHub was created in June 2022 and is licensed under Apache-2.0. As of the available data, the repository has 410 stars and 72 forks, with the last code push recorded in June 2024. The project also publishes standalone open-source packages including speech-recorder (a native npm module for microphone access and voice activity detection) and serenade-driver (a native npm module for OS-level keystroke and mouse hooks). Community support is provided through a Discord server linked from the project's documentation.