KDE is building a second terminal emulator — and that decision is the story, not the app. According to Phoronix, developers have chosen to start from a clean slate with a project called Kwi, built on the Ghostty terminal library, rather than continue refactoring Konsole, the long-standing terminal that ships with Plasma. The report carries a headline claim of performance gains between 3x and 35x.
Those numbers warrant caution. Per the report, they come from the developer's own test conditions, and the gap between a threefold and a thirty-fivefold result says more about workload variance than about a uniform speedup. Kwi is also an early developer preview: no stable release, no feature parity with Konsole, no independent verification of the benchmarks. Read the figures as a directional signal, not a settled comparison.
The architectural case
The project's logic rests on architecture rather than arithmetic. Kwi borrows Ghostty's modern rendering and I/O core — the parts of a terminal that determine how responsive it feels under heavy output, fast scrolling, and multiplexed sessions — and wraps it in a KDE-native interface. The result is a Plasma application that presents itself as a first-party KDE offering while running on a foundation designed for contemporary hardware acceleration.
Konsole, by contrast, carries decades of accumulated code. It can be modernised in place, but every improvement is hemmed in by design decisions made when the terminal landscape looked very different. Running a second terminal in parallel lets KDE experiment without endangering the tool many Plasma users depend on daily.
That trade-off is not unique to KDE. The terminal world has been grappling with it for years: kitty and Alacritty staked their reputations on GPU-accelerated rendering pipelines that, their authors argue, deliver noticeably smoother behaviour than legacy terminals — particularly with large scrollbacks or Unicode-heavy output. Kwi brings that same argument inside the KDE fold, with KDE styling, settings integration, and conventions layered on top.
The same question recurs across the open-source ecosystem: when a long-shipping component carries architectural debt, should it be refactored in place, or should a parallel replacement be built while the original stays in service? KDE's answer — develop Kwi alongside Konsole rather than retire it — is a pattern downstream distributions and other desktop projects will recognise as a familiar compromise.
Current status and what it means
Kwi today is an early build — a developer preview, not a shipped, supported application. Users should expect rough edges and missing functionality relative to Konsole, and should treat it as a demonstration of direction rather than a drop-in replacement.
Should the project mature, Plasma users gain something rare on the Linux desktop: a choice between two terminals from the same project, one battle-tested and one built for modern performance. Even if Kwi never graduates beyond proof of concept, it raises a question KDE will eventually have to answer — whether Konsole's architecture is worth revisiting, or whether the future lies in composing KDE experiences on shared, cross-platform engines.
For now, the benchmarks remain unverified and no release date exists. What is clear is that KDE is willing to question its own assumptions — and that, more than any multiplier, is the news worth watching.
KDE 正在開發第二款終端模擬器——而這個決定本身才是新聞焦點，並非該應用程式。據 Phoronix 報導，開發人員選擇另起爐灶，開展名為 Kwi 的新項目，以 Ghostty 終端函式庫為基礎構建，而非繼續重構隨 Plasma 出貨的老牌終端 Konsole。該報導的標題聲稱效能提升介乎 3 倍至 35 倍之間。
這些數字值得審慎看待。據報導，它們來自開發者自身的測試條件，而三倍與三十五倍之間的差距，反映的與其說是普遍的加速，不如說是工作負載的差異。Kwi 目前亦只是早期開發者預覽版：沒有穩定版本，功能未及 Konsole，基準測試亦未經獨立驗證。讀者應將這些數據視為方向性指標，而非已成定論的比較。
架構上的理據
項目的邏輯建基於架構而非數字。Kwi 借用了 Ghostty 的現代化渲染及 I/O 核心——即決定終端在大量輸出、快速捲動及多工分隔畫面下回應表現的部分——並在外層包覆 KDE 原生介面。成果是一款 Plasma 應用程式，對外定位為 KDE 第一方應用程式，底層卻運行於專為當代硬件加速而設計的基礎之上。
相比之下，Konsole 承載著數十年累積的代碼。它固然可以就地現代化，但每一項改進都受制於終端生態截然不同時所作出的設計決定。平行運行第二款終端，讓 KDE 得以在不危及眾多 Plasma 日常用戶所依賴工具的前提下進行實驗。
這種取捨並非 KDE 獨有。終端領域多年來一直在應對同一課題：kitty 和 Alacritty 將聲譽押注於 GPU 加速渲染流水線，其作者認為它們較傳統終端帶來明顯更流暢的表現——尤其在大量捲動緩衝或 Unicode 密集輸出時。Kwi 將同一論述引入 KDE 陣營，並在其上疊加 KDE 樣式、設定整合及使用慣例。
同一問題在開放源碼生態中反覆出現：一個推出已久的元件若累積了架構負債，究竟應就地重構，還是另建平行替代方案並保留原有元件繼續服役？KDE 的答案——與 Konsole 並行開發 Kwi 而非將其退役——是下游發行版及其他桌面項目都會視為熟悉的折衷做法。
目前進度及其意義
Kwi 目前只是早期版本——屬開發者預覽，而非已推出並獲支援的應用程式。用戶應預期較 Konsole 為多的粗糙之處及功能缺失，並應視之為方向展示，而非可直接替換的方案。
倘若項目日後成熟，Plasma 用戶將可獲得 Linux 桌面上罕見的選擇：同一項目提供的兩款終端，一款久經考驗，一款為現代效能而建。即使 Kwi 最終未能超越概念驗證階段，它亦拋出了一個 KDE 終須回答的問題——Konsole 的架構是否值得重新檢視，還是未來在於以共享的跨平台引擎來組合 KDE 體驗。
目前而言，基準測試仍未經驗證，亦不存在發佈日期。可以確定的是，KDE 願意質疑自身的既定假設——而這一點，比任何倍數都更是值得關注的新聞。