CutScript

CutScript is a desktop audio and video editor for podcasters and creators who prefer editing speech through text. Deleting a word from the transcript cuts the matching section of the recording, making it a local-first Descript alternative.

• Transcript editing: WhisperX aligns the transcript to individual words, speaker diarization distinguishes speakers, and a waveform timeline lets you review edits with undo and redo.
• AI cuts and clips: Detect filler words and suggest short clips from longer recordings using Ollama locally or OpenAI and Claude as cloud providers.
• Audio cleanup: Reduce background noise with DeepFilterNet before exporting the edited recording.
• Captions and export: Create word-level captions as SRT, VTT or ASS files, burn captions into video, and export through FFmpeg using stream copy or re-encoding up to 4K.

The desktop app and its Python backend run on your machine. Ollama provides the local option for AI features; choosing OpenAI or Claude sends those AI requests to the selected cloud provider. CutScript is open source under the MIT License and requires Python, Node.js and FFmpeg.