A Zig toolkit for building desktop apps with native UI drawn on the GPU, or with a web page in the system webview.
longbridge/gpui-fast
A performance-focused fork of Zed's GPUI Rust UI framework with incremental rendering and native window composition.
- Stars
- 411
- Forks
- 52
- Open issues
- 2
- Last commit
- 13 hours ago
What is it?
GPUI Fast is a fork of Zed's GPUI, the Rust UI framework, extracted from the Zed monorepo as a standalone workspace. It targets rendering cost: the retained mode tracks rendering dependencies so unchanged views, layout nodes and text measurements are reused between frames rather than rebuilt.
The fork also adds scroll layers that reuse cached GPU tiles and list rows while scrolling on macOS, Linux and Windows. An opt-in window composition mode embeds native content such as WKWebView, WebView2 or WebKitGTK views while keeping GPUI menus, popovers and dialogs rendered above them.
The project is published on crates.io as gpui-fast and gpui-fast-platform, and is marked experimental.
Key features
- 01Retained mode reuses unchanged views, layout nodes and text measurements between frames.
- 02Scroll layers reuse cached GPU tiles and list rows while scrolling.
- 03Window composition embeds native web views while GPUI overlays stay above them.
- 04Supports macOS, Windows, and Linux on both Wayland and X11.
- 05Published to crates.io and aliasable as a drop-in replacement for the gpui crate.
Use cases
- Building a desktop application in Rust that needs GPUI with lower per-frame CPU cost.
- Embedding a native web view inside a GPUI window while keeping GPUI dialogs above it.
- Evaluating incremental rendering for apps with largely static windows or long scrolling lists.
Quick start
- 1
Add the published crates to your Cargo.toml
- 2
Run the native webview example on macOS, Windows or Linux
cargo run -p gpui_perf --example native_webview
Things to know
- The project remains experimental.
- Linux platform features are required, and the library selects different backends on each platform.
- Apache-2.0 fork; the README notes it is modified from Zed's GPUI.
In collections
Topics
FAQ
- What is gpui-fast?
- A performance-focused fork of Zed's GPUI Rust UI framework with incremental rendering and native window composition. GPUI Fast is a fork of Zed's GPUI, the Rust UI framework, extracted from the Zed monorepo as a standalone workspace. It targets rendering cost: the retained mode tracks rendering dependencies so unchanged views, layout nodes and text measurements are reused between frames rather than rebuilt.
- Is gpui-fast free to use? What license does it have?
- gpui-fast is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is gpui-fast still maintained?
- Actively maintained. The last commit to gpui-fast was on Nov 6, 2026.
- What can I use gpui-fast for?
- Building a desktop application in Rust that needs GPUI with lower per-frame CPU cost. Embedding a native web view inside a GPUI window while keeping GPUI dialogs above it. Evaluating incremental rendering for apps with largely static windows or long scrolling lists.
- How do I install gpui-fast?
- Add the published crates to your Cargo.toml Run the native webview example on macOS, Windows or Linux: cargo run -p gpui_perf --example native_webview
- Where can I get gpui-fast?
- The source code and installation instructions are on GitHub: https://github.com/longbridge/gpui-fast
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A curated awesome-list of desktop apps, libraries and tools built with or for the GPUI Rust UI framework.
A native, GPU-rendered Rust UI framework for macOS, Windows, Linux and the web, with its own renderer, text, layout and springs.
A cross-platform Rust UI framework that ports Avalonia's architecture, with styled properties, data binding, XAML markup and multiple themes.
Open-source Python framework for building GUI apps that run on desktop, mobile and embedded platforms from one codebase.
A cross-platform C++17 UI framework with GLFW/SDL2 window backends and OpenGL/Vulkan rendering, built around a declarative DSL.