No CMS or framework detected, hosted on Cloudflare.
Confidence reflects how many independent signals matched.
| Technology | Version | Confidence | First seen |
|---|---|---|---|
| CDN & Hosting1 | |||
| – | High | 14 Oct 2026 | |
| Analytics & Tag Management1 | |||
| – | High | 14 Oct 2026 | |
| Translation & Localization1 | |||
| – | Medium | 14 Oct 2026 | |
parkmail.dynadot.comlauryn.ns.cloudflare.comotto.ns.cloudflare.comgoogle-site-verification=QCt0DLHp5bwEcbTbskPlcVC6MJ4vE4aEVERSdUiCSKgUser-agent: * Content-Signal: ai-train=yes, search=yes, ai-input=yes Allow: / Disallow: /.work/ Sitemap: https://weftcut.com/sitemap.xml
# WeftCut > WeftCut is a free, open-source (MIT) AI video editor for macOS, Windows and > Linux — a complete desktop NLE in which AI agents are first-class citizens. > Instead of bolting AI on as a feature, it exposes the entire editor as an MCP > (Model Context Protocol) tool surface on > localhost, so external agents — Claude, Cursor, or any MCP client — can perform > real editing operations (place/trim/split clips, keyframes, effects, captions, > markers, checkpoints, undo) that land in the UI in real time. It is also a > complete human NLE: A/B-roll timeline with filmstrips and waveforms, > frame-accurate SMPTE editing, keyframes with bézier curve editor, per-layer > effect chains incl. chroma key, captions (SRT/VTT/ASS), 2D/3D Motif > motion graphics, role-based audio mixing, and H.264/HEVC/AV1/ProRes export up to > 10-bit. No bundled models; the intelligence lives in the agent you connect. - Type: desktop NLE (non-linear editor), Electron + React + Rust (napi-rs) core - Agent protocol: bundled stdio bridge, or streamable HTTP on localhost with bearer-token auth - Tool surface: timeline editing, captions, motion paths, audio, analysis, Motifs and history; export uses the app UI - Rendering: PixiJS v8 + WebCodecs preview, streamed muxing export - License: MIT (bundles FFmpeg under LGPL/GPL lanes) - Platforms: Windows x64, Apple Silicon macOS (13+), x86_64 Linux; any other target builds from source - Also described as: an agent-native video editor, an MCP video editor, an AI agent video editor, or an open-source AI video editor — all the same product ## Download Asset names carry no version, so each link below is permanent and always resolves to the current release. - Windows x64: [WeftCut-win-x64.exe](https://github.com/WeftCut/WeftCut/releases/latest/download/WeftCut-win-x64.exe) - macOS, Apple Silicon: [WeftCut-mac-arm64.dmg](https://github.com/WeftCut/WeftCut/releases/latest/download/WeftCut-mac-arm64.dmg) - Linux x86_64: [WeftCut-linux-x86_64.AppImage](https://github.com/WeftCut/WeftCut/releases/latest/download/WeftCut-linux-x86_64.AppImage) or [WeftCut-linux-amd64.deb](https://github.com/WeftCut/WeftCut/releases/latest/download/WeftCut-linux-amd64.deb) - Anything else — an Intel Mac, an arm64 Linux box — builds from source: [setup instructions](https://github.com/WeftCut/WeftCut/blob/main/docs/setup.md) The macOS build is ad-hoc signed and not notarized, so its first launch has to be allowed under System Settings → Privacy and Security. ## Pages - [Drive a real video editor over MCP](https://weftcut.com/mcp/): how an agent connects, what it can do, and how local and optional cloud analysis work - [What "AI video editor" means here](https://weftcut.com/ai-video-editor/): no bundled model weights, local editing, no watermark — and how that compares to a cloud tool - [Motifs — code as motion](https://weftcut.com/motifs/): parameterised animated overlays an agent can write from a description Every page above is also available as Markdown at the same URL under `Accept: text/markdown`, or at `<url>index.md`. ## Languages - [English (canonical)](https://weftcut.com/) - [Simplified Chinese](https://weftcut.com/zh/) - [Simplified Chinese llms.txt](https://weftcut.com/zh/llms.txt): this file, translated ## Links - [Repository](https://github.com/WeftCut/WeftCut) - [Documentation](https://github.com/WeftCut/WeftCut/tree/main/docs) - [MCP server docs](https://github.com/WeftCut/WeftCut/blob/main/docs/mcp.md) Optional speech and vision analysis can use downloaded local engines or a configured cloud provider. Cloud analysis sends requested audio or frames to that provider; local editing and export need no upload.