The Editor War Nobody Expected
VS Code has dominated the developer tools market for years, and Cursor carved out a profitable niche by layering AI on top of it. But a third player is now pulling developers away from both – not with AI gimmicks or venture-backed marketing, but with raw performance. Zed, the code editor built by the team that created Atom, is gaining a dedicated following among developers who have grown frustrated with the bloat, latency, and plugin chaos that comes with VS Code’s flexibility.
Zed is not trying to be everything to everyone. It is written in Rust, starts in milliseconds, and handles large codebases without the memory overhead that plagues Electron-based editors like VS Code. That is a specific pitch aimed at a specific group – the power users who know exactly what their tools are doing to their CPU and are tired of watching a text editor eat 2GB of RAM while they type.
What Zed Actually Does Differently
The performance gap between Zed and VS Code is not subtle. VS Code runs on Electron, which wraps a Chromium browser around the editor interface. That architecture made VS Code easy to build and extend, but it also means the editor carries significant overhead at all times. Zed, built natively in Rust with its own GPU-accelerated rendering layer called GPUI, skips all of that. The result is an editor that feels immediate in a way Electron apps simply cannot match.
For developers working in large monorepos – the kind common at mid-size startups and mature engineering teams – that difference is felt constantly. File navigation, symbol search, and syntax highlighting in Zed stay fast as file counts grow. In VS Code, those same operations slow noticeably once a project passes a certain scale. This is not a configuration problem that plugins can fix. It is an architectural limitation baked into how the editor was built, and Zed’s team has made that contrast the center of their positioning.
Zed also ships with built-in multiplayer editing, a feature that lets multiple developers edit the same file simultaneously with presence indicators – something VS Code has never natively supported and relies on third-party extensions to approximate. The implementation in Zed feels closer to a Google Docs-style real-time session than the clunky shared terminal hacks developers typically use for pair programming. It is a genuinely useful feature for teams doing code review together or onboarding junior engineers.
The AI integration Zed ships is deliberately restrained. Rather than building an AI-first product like Cursor did, Zed treats AI as one feature among many and lets developers configure it to point at different model backends, including local models. Developers who want AI heavily integrated into their workflow will likely still prefer Cursor. But developers who want AI available without having it reshape the entire editing experience are finding Zed’s approach more tolerable.
Why VS Code’s Extension Ecosystem Is Both a Strength and a Trap
VS Code’s greatest competitive advantage is also what drives its most sophisticated users away. The extension marketplace has over 50,000 extensions, covering every language, framework, and workflow imaginable. That breadth is why VS Code became the default editor for so many developers. But extensions add overhead, introduce conflicts, and require maintenance. A VS Code installation that has grown organically over years often becomes a liability – slow, unpredictable, and full of half-working plugins from abandoned repositories.
Zed has very few extensions by comparison. It supports a narrower set of languages natively, and its extension ecosystem is still in early development. For most developers, this is a dealbreaker. But for the subset who have spent years trying to make VS Code fast and finally given up, the simplicity is the point. A clean, fast editor with great defaults is more useful to them than an endlessly customizable one that requires constant maintenance to stay functional.
Cursor Is the Immediate Competitor, But Zed Has Different Ambitions
Cursor gets most of the attention in developer tool discussions right now because it is explicitly an AI coding assistant wrapped in an editor. It raised significant funding and attracted developers who want AI suggestions deeply embedded in their editing flow. Zed’s pitch is different enough that the two are not always competing for the same user at the same moment – Cursor users tend to be optimizing for AI output, while Zed users tend to be optimizing for the editing experience itself.
That said, Zed is adding AI features steadily. Its current AI capabilities include inline code generation, a chat panel, and context-aware completions. As those features mature, the distinction between Zed and Cursor narrows, and the remaining differentiator – Zed’s native performance – becomes the deciding factor. Cursor is still built on VS Code’s foundation, which means it inherits VS Code’s architectural constraints along with its extension compatibility. Zed has no such baggage.
The business model Zed is pursuing is still taking shape. The editor is open source, and the team has indicated that collaboration features and hosted services will be where they eventually charge. That mirrors how other developer tools have monetized – free for individuals, paid for teams – and it is a model the market understands. What makes it interesting is that Zed is betting the team coordination layer becomes valuable enough to sustain the company, while the editor itself stays free.
The Power User Migration Problem
Power users in developer tools matter disproportionately. They write the blog posts, make the YouTube tutorials, and set the defaults at companies where they have influence over tooling decisions. When a senior engineer at a startup switches to a new editor and talks about it publicly, junior developers on the same team notice. That is how VS Code itself spread – not through enterprise sales, but through individual developer advocacy that accumulated over years into a near-monopoly.
Zed is now benefiting from a version of that same dynamic. Developer forums and social channels have seen a steady increase in posts from longtime VS Code users describing the switch, with most citing performance as the primary reason. The feedback loop of senior developers advocating for Zed is early but real, and the editors that win developer mindshare this way tend to keep it. VS Code’s share of that vocal, influential layer of the developer population is smaller than it was two years ago, and Zed is where a meaningful portion of those developers went.
The question that remains unresolved is whether Zed can hold those users once its extension ecosystem is more developed and the inevitable complexity begins to set in. Right now, Zed’s simplicity is partly a feature and partly just the early stage of a product that has not yet accumulated the weight of maturity. If it follows the trajectory of every other editor that started fast and focused, the real test will come when the user base is large enough to demand the very things that made VS Code slow.
