Video editor
A real timeline, not a trimmer. Zoom, speed ramps, cursor packs, backgrounds and the AI Speech Enhancer. What you see is what renders.
Pixel-perfect screenshots, scrolling capture, on-device OCR, draw on screen, screen recording, and a real multi-track video editor.
Features
Every feature is a switch. Turn off what you don’t use and it disappears.
A real timeline, not a trimmer. Zoom, speed ramps, cursor packs, backgrounds and the AI Speech Enhancer. What you see is what renders.
Screen, webcam, mic and system audio as separate tracks. Cursor and keys recorded as data, so you can restyle them later.
Ink over any app, then press Esc and keep working underneath it. Live keystrokes included.
Press PrtScn. Zoom to 256×, click a window, get the window. Every button is one letter.
Copy text from anything on screen. It runs on your machine and shows you what it read.
The whole page as one tall image. Sticky headers appear once.
Curved arrows, numbered steps that point, redaction that never blurs your notes. One key per tool.
Share any rectangle in any meeting app. Blur or black it out without stopping.
To your own S3, Azure, R2 or B2. The link is on your clipboard before the upload finishes.
Video Editing
Zoom, speed, cursor, background and audio cleanup on a real multi-track timeline, then one click to export a file that’s ready to post or send.
Ctrl+K splits every grouped row. The gap menu puts it back.
Zoom where attention should land. Skip the waiting at up to 10×.
Restyle the cursor and show shortcuts, after recording.
Backgrounds, rounded corners and shadow in one click.
Separate tracks. Screen, webcam, mic and system audio.
On-device AI. Speech Enhancer and background removal on your GPU.
One-click export. Share, Best quality or Small file.
Live Presenting
Walkthroughs, onboarding and code reviews that feel like you’re sitting next to them. Every shortcut you press shows up on screen, so nobody has to ask what you just did.
Every tool is one keypress away, so you never break stride mid-sentence.
Esc: the ink stays put while you scroll underneath it.
Live keystrokes, from the same renderer as the editor.
Freeze a moving screen and draw on the still.
Draw, then Esc. Your drawings stay on screen while you click, scroll and type underneath.
Eraser and undo across every monitor. One history for every display.
Freeze the frame or keep it live. Draw on a still, or right over the moving screen.
Screenshots
Snap to a whole window in one click, or zoom in and crop to the exact pixel. Then take it anywhere in Clowd with a single key: annotate it, record it, copy it or share it.
Click a window, get the window. Even one hidden behind another.
Scroll to zoom, then place the edge on the exact pixel.
One toolbar, every next step. Each button is one letter.
Two keys from screen to clipboard.
Press H on any pixel for its color.
Owns PrtScn. Even on Windows 11.
Sees behind windows. Grabs a window hidden under others.
Keyboard first. Enter edits, Esc backs out.
Privacy
Everything Clowd does by default happens on your computer and stays there.
x64
Windows 10+
ARM64
Windows 10+
Apple Silicon
macOS 13+
Intel
macOS 13+
FAQ
Yes. There is no trial, no paid tier and no watermark. Every feature on this site is in the download.
Clowd is open source and has been in development since 2015. If you upload to your own storage, your provider bills you for that storage, not us.
Yes. Clowd runs natively on Macs with Apple Silicon and Intel, on macOS 13 or later, with the same capture overlay, editors, recorder, Draw on Screen and Share Region as on Windows.
A few things differ. Text recognition and the AI video effects need Apple Silicon. The "Upload with Clowd" right-click menu is a Windows 11 feature. The default hotkeys use PrtScn, so on a Mac keyboard you will want to rebind them. Scrolling capture asks for Accessibility permission.
Requires Windows or a Mac with Apple Silicon. Windows 11. Default hotkeys use PrtScn. Rebind them in Settings → Hotkeys on a Mac keyboard. Scrolls whole windows, not panes inside them. Needs Accessibility permission on macOS.
Only where you send them. The Upload button on the capture toolbar is off until you turn it on, and the whole Uploads feature can be switched off.
Images, videos and files go to a destination you set up: Amazon S3 or any S3-compatible service, Azure Blob Storage, Cloudflare R2, Backblaze B2, or a zero-config host such as catbox. Text goes to clwd.app pastes by default.
With accelerated uploads, the link is on your clipboard the moment the upload starts. The file streams through the clwd.app relay into your bucket, and once it is there the link points straight at your storage. Nothing stays in the middle.
Image and video uploads need a storage destination; text pastes work out of the box.
No. Text recognition runs on your computer, in Clowd's own on-device engine. There is no cloud service, no API key and no language pack to download.
The Search button is different and does what it says: it sends the selection to a reverse image search in your browser, and only when you press it.
Requires Windows or a Mac with Apple Silicon. Latin, Greek, Chinese and Japanese kana. Cyrillic and Korean aren’t supported yet.
Both. Clowd ships native builds for Windows x64, Windows ARM64, Macs with Apple Silicon and Intel Macs, each as an installer or a portable zip.
Windows builds are code-signed and Mac builds are notarized. Text recognition and the AI effects run on Windows (x64 and ARM64) and on Apple Silicon, but not on Intel Macs.
Requires Windows or a Mac with Apple Silicon.
Clowd’s capture overlay, recorder and editors are built on each platform’s own graphics and capture APIs, so Linux needs its own port, and it isn’t ready. There is no Linux download today.
If you are on Linux now, Flameshot and OBS Studio are good open source choices. To follow Clowd’s progress, watch the repository on GitHub.
Not available yet.
Clowd updates itself from GitHub Releases. It checks every few hours, downloads in the background, and restarts to apply an update only after 10 minutes of inactivity, never during a capture, recording or upload.
You can change how often it checks, turn off automatic restarts, or opt in to experimental builds in Settings.
MIT. You can read, change and redistribute it, including for commercial use.
The one exception is the on-device AI sidecar that runs text recognition and the AI video effects. It is GPL-3.0, because the video-matting model it bundles is GPL-3.0.
MIT, except the on-device AI sidecar (GPL-3.0).
Free. Open source.Download ClowdView on GitHub