GitHub's New Copilot App Lets AI Help You Code Without Opening a Text Editor
A New Way to Write Code
GitHub has released a test version of a new desktop application that brings AI coding assistance into its own dedicated window, rather than tucking it inside existing code editors. The app is currently available to paying Copilot customers, with other users able to request early access through a waitlist. You can learn more about access here.
Until now, GitHub's Copilot AI tool has worked as an add-on to existing software that developers already use—tools like Visual Studio Code, JetBrains IDEs, and text editors. The new desktop app takes a different approach: it builds its entire interface around having an AI assistant take on more complex tasks, rather than just suggesting code snippets.
Keeping Your Work in One Place
The application is designed around a concept called "sessions." Think of a session as a conversation thread between you and the AI that remembers everything relevant to a specific project task. You can start a session from a to-do list, a code review comment, a simple request, or by continuing an earlier conversation.
The app includes a dashboard called "My Work" that shows what you are currently working on across all your projects. Normally, developers juggle multiple windows and browser tabs to see everything happening at once—their task list in one place, their code in another, their pull requests somewhere else. This new app tries to gather all of that into one view.
GitHub is marketing the application with a phrase: "Direct agents from issue to merge." What that means in plain language is that the AI can help guide your work all the way from identifying a problem that needs fixing, through writing code, all the way to merging that code into your project.
Who Can Use It and Why
The desktop application is a native program running directly on your computer, rather than a website you visit in a browser. GitHub has not shared detailed technical information about how it works during this early testing phase.
GitHub is making the app available first to its most committed paying customers. This suggests the company sees it as an advanced feature, not a replacement for its existing Copilot tools built into code editors. People who do not yet have a paid Copilot account can still request early access.
Why This Matters
The broader context here is that AI tools across the technology industry are shifting from doing single, isolated tasks to handling entire workflows—a series of connected steps. GitHub's bet is that developers will switch to new software when the efficiency gains are significant enough.
This is not the first time we have seen this happen. In the 1990s and early 2000s, developers were attached to using command-line tools for managing code versions. Over time, they gradually switched to separate applications with visual interfaces—programs like SourceTree and GitKraken—because those tools made complex code management much easier to see and understand.
GitHub currently has about 1.8 million developers using Copilot. The new desktop app is designed to deepen that relationship by expanding what the AI can do—moving beyond suggesting snippets of code to helping orchestrate entire development workflows.
The real test will be whether developers feel the productivity gains justify learning new software. People tend to stick with familiar tools unless there is a clear and compelling reason to switch. The early testing phase will tell GitHub whether developers see the benefit as worth the change.
For teams already using GitHub for their projects, the desktop app offers a path to more integrated workflows without requiring them to change how they store or share their code. The app is built to keep track of context—remembering details from task lists, code reviews, and automated checks—which normally gets scattered across different tools and windows in distributed teams.
The question now is whether GitHub's approach will succeed, or whether developers will prefer to keep using their familiar code editors with AI assistance added as a bonus feature rather than switching to a completely new interface built around AI collaboration.
