Cephalopod Studio

Director's Cut

Raw footage in.
Finished video out.

AI-powered video editing from the command line. One prompt, one command, one finished video with narration and subtitles.

$ directorscut edit \
  -p "Create a 30-second product demo with upbeat narration" \
  -f ./raw-footage/ \
  -o product-demo.mp4 \
  --narrate --subtitle

Analyzing 47 clips... done
Generating edit decisions... 12 scenes selected
Writing narration script... 84 words
Synthesizing voiceover... done
Rendering timeline... done

Saved to product-demo.mp4 (0:32)

Condensed for demonstration

Everything you need, nothing you don't

Prompt-Driven Editing

Describe the video you want in plain English. The AI selects clips, sequences them, adds transitions, and renders your timeline.

🎤

AI Narration

Auto-generated voiceover scripts timed to your video. Use cloud TTS or clone your own voice locally with Chatterbox — your voice data never leaves your machine.

Animated Subtitles

TikTok-style word-by-word subtitles with precise timing. Multiple styles to match your content.

🔍

Transparent Decisions

Every edit decision is saved as readable, editable JSON. Tweak the timeline by hand and re-render instantly — no AI call needed.

🚀

Batch Production

Run the same footage with 10 different prompts to get 10 unique videos. A/B test edits at scale without touching a timeline.

🔄

NLE Export

Export rough cuts as OpenTimelineIO for finishing in Premiere Pro, DaVinci Resolve, or Final Cut Pro.

How it works

1

Point it at your footage

Pass a folder of raw clips. Director's Cut detects scenes, transcribes speech, and builds a visual understanding of every shot.

2

Describe what you want

Write a prompt: "30-second product demo with narration" or "highlight reel of the best action shots." The AI makes the editorial decisions.

3

Get a finished video

Out comes a rendered MP4 with transitions, narration, and subtitles. Publish directly to YouTube or TikTok, or export to your NLE for final polish.

Made with Director's Cut

Real videos generated from raw footage with a single command.

Image-to-3D AI Generation

Generating 3D models from images using AI, narrated and edited automatically.

Multi-Object Comparison

Multiple scanned objects compared in a single narrated video.

Getting Started Tutorial

An auto-generated how-to tutorial from raw screen recordings.

Built for people who'd rather ship than scrub timelines

3 free video generations to try before you buy.

Content Creators

Batch-produce TikToks, Reels, and Shorts from a single shoot. Run the same footage 10 times with different prompts, pick the best one.

Developers & PMs

Turn screen recordings into polished product demos. No timeline scrubbing. No drag-and-drop. Just a prompt and an MP4.

Agencies & Marketing

Generate rough cuts in seconds, then hand off to your editor via OTIO export. A/B test video edits without multiplying production time.

Educators

Auto-narrate lecture footage, add subtitles for accessibility, and publish directly — all from one command.

$29
One-time purchase. Yours forever.
No subscription. No per-minute pricing. All future updates included.
  • 3 free video generations to try first
  • Unlimited use after purchase
  • AI narration & animated subtitles
  • Local voice cloning (on-device, private)
  • YouTube & TikTok publishing
  • OpenTimelineIO export to pro NLEs
  • Works fully offline on Apple Silicon
Get a License Key

Requirements

Python 3.11+ (3.14 recommended)
FFmpeg (free, open source)
Gemini API key (free tier available)
macOS or Linux