Cutted is an AI-powered audio editor you control with natural language.
- Cut, trim, or adjust volume by typing what you want
- Detect and remove silence or loud parts automatically
- Transcribe audio (with Whisper)
- Undo edits, export as MP3 or WAV
- Make sure your Python interpreter has Tkinter support
- Make sure FFmpeg is installed.
- Install Cutted:
With Whisper support:
pip install cutted
pip install cutted[whisper]
python -m cutted
Load an audio file, enter your command in plain English, and export your result.
You can also enable or disable chat history for smarter follow-up actions in the settings menu (gear icon).
-
Cut a segment:
cut from 10 to 20
Removes audio between 10s and 20s.
-
Remove multiple segments:
cut from 5 to 7 and from 15 to 18
-
Change volume:
make 0 to 5 seconds twice as loud
or
reduce volume from 10 to 12 by half
-
Remove silence:
remove all silent parts
Note: Removing silence is experimental and can be buggy or produce bad results in some cases.
-
Undo last edit:
Click the "Undo" button. -
Export your result:
Click the "Export" button and choose MP3 or WAV.
If you install with Whisper support and enable "Send transcript to Gemini" in the UI, you can use commands like:
- Cut by word or sentence:
or
cut when I say the word hello
cut the sentence "welcome to the show"
This uses Whisper to transcribe your audio and allows precise editing based on spoken words or phrases.
If you enable "Send audio to Gemini" (checkbox), you can also use word-based commands, but accuracy is much lower compared to using Whisper.
Set your Gemini API key as an environment variable before running Cutted:
Linux/macOS:
export GEMINI_API_KEY="your-gemini-api-key"
Windows (Command Prompt):
set GEMINI_API_KEY=your-gemini-api-key
Windows (PowerShell):
$env:GEMINI_API_KEY="your-gemini-api-key"