ClipTool

Voice-activated screen clipper for Windows. Say "computer clip that" and save the last few minutes of your screen + audio instantly.

Features

• Voice activation — Say "computer clip that" to save a clip hands-free
• GPU-accelerated capture — Uses DXGI (same as OBS) for zero-lag recording
• Variable frame rate — Perfect audio/video sync at any resolution and FPS
• Dual audio — Records microphone + desktop audio simultaneously with per-source volume control
• High resolution — Supports 720p, 1080p, 1440p, and native monitor resolution
• Configurable FPS — 30, 60, or 120fps capture
• System tray — Runs quietly in the background, always ready
• Hotkey — Ctrl+Shift+F9 to clip without speaking
• Persistent settings — All preferences saved to config.json

Prerequisites

Requirement	Notes
Windows 10/11	Required for DXGI capture
Python 3.10+	Download Python
FFmpeg	Must be in PATH or installed via WinGet
Microphone	For voice commands and mic recording
VB-Cable (optional)	For desktop audio capture (download)

Installation

Option A: Standalone EXE (Windows)

Download the latest release from the GitHub Releases page. No installation needed — just run the .exe file.

Option B: Install via pip

pip install clip-tool

Then run:

clip-tool

Option C: Run from source

1. Install Python

Download from python.org. Make sure to check "Add Python to PATH" during installation.

2. Install FFmpeg

Option A — WinGet (recommended):

winget install Gyan.FFmpeg

Option B — Manual:

1. Download from ffmpeg.org
2. Extract to C:\ffmpeg\
3. Add C:\ffmpeg\bin to your system PATH

3. Install VB-Cable (optional, for desktop audio)

Download from vb-audio.com/Cable and run the installer. This creates a virtual audio device that lets ClipTool record your desktop audio (game sounds, music, etc.).

4. Clone and install dependencies

git clone https://github.com/Sexigt/clip-tool.git
cd clip-tool
pip install -r requirements.txt

5. Run

python main.py

The app starts minimized in the system tray. The main window appears after a few seconds.

Usage

Voice Commands

Speak naturally after the wake word "computer":

Say	Action
"computer clip that"	Save the last clip (uses default duration)
"computer clip last 30 seconds"	Save the last 30 seconds
"computer clip last 2 minutes"	Save the last 2 minutes

The app plays a clipped.mp3 sound when a clip starts saving.

Manual Controls

Button	Action
Clip That	Save using the default duration
1min / 2min / 5min	Save that many seconds from the buffer
Open Folder	Open the clips folder in Explorer

Hotkey

Press Ctrl+Shift+F9 anywhere to save a clip (uses default duration).

System Tray

Right-click the tray icon for quick access:

• Show — Bring up the main window
• Clip 1min / 5min — Quick save
• Open Folder — Browse saved clips
• Quit — Exit the app

Settings

Capture

Setting	Options	Default	Notes
Resolution	720p / 1080p / 1440p / Native	1080p	Native uses your monitor's full resolution
FPS	30 / 60 / 120	60	Higher = smoother but more RAM
Quality	70 / 85 / 95	95	JPEG quality during capture. Higher = sharper

Audio

Setting	Range	Notes
Mic volume	0–200%	Volume applied to mic audio in saved clip
Desktop volume	0–200%	Volume applied to desktop audio in saved clip
Record Mic	On/Off	Toggle mic recording
Record Desktop	On/Off	Toggle desktop audio recording
Default duration	30s / 1min / 2min / 5min	Duration used by "computer clip that"

All settings are saved to ~/clip-tool/clips/config.json and restored on next launch.

File Structure

clip-tool/
├── main.py              # Main application
├── requirements.txt     # Python dependencies
├── audio/               # Sound effects
│   ├── clipped.mp3      # Played when clipping
│   ├── sound-on.mp3     # Played when toggling recording on
│   └── sound-off.mp3    # Played when toggling recording off
└── clips/               # Saved clips (created on first run)
    ├── clip_*.mp4       # Your saved clips
    ├── config.json      # Saved settings
    ├── _temp/           # Temporary files (auto-cleaned)
    └── cliptool.log     # Error log

Memory Usage

ClipTool keeps a ring buffer of captured frames in RAM. Usage depends on resolution and FPS:

Resolution	FPS	Buffer	~RAM
720p	30	5 min	~0.7 GB
1080p	60	5 min	~3.6 GB
1440p	60	5 min	~6.3 GB
Native 2K	120	5 min	~12 GB

Lower the FPS or buffer duration if you experience memory pressure.

Troubleshooting

"ffmpeg not found"

• Make sure FFmpeg is installed and in your PATH
• Restart your terminal after installing

Voice commands not detected

• Check that your microphone is selected in the Mic dropdown
• Speak clearly after "computer" — the wake word needs to be distinct
• Check cliptool.log in the clips folder for errors

Desktop audio not recording

• Install VB-Cable
• Set VB-Cable as your default playback device, or use it as a monitoring device
• Select the loopback device in the Desktop dropdown

Clip is blurry

• Increase the Quality setting (95 recommended)
• Make sure Resolution is set to your monitor's native resolution

Audio out of sync

• The app uses variable frame rate encoding for proper sync
• If sync is off, try lowering the FPS setting

Built With

• bettercam — DXGI screen capture
• faster-whisper — Speech recognition
• customtkinter — UI framework
• FFmpeg — Video/audio encoding
• pyaudiowpatch — Audio capture

License

MIT

Star History