The video editor that edits for you.
Tell it what you want. It cuts the timeline, writes the captions, generates the footage, and builds the graphics, on your Mac, with the AI provider you choose, on footage that never leaves your machine.
No subscription. No account. Bring your own AI keys, or run a free local model.
Own your tools
Software you buy, not software you rent.
Every serious creative tool is drifting to subscriptions and clouds. This one goes the other way: a native Mac app you pay for once and own outright. The source is right there if you'd rather compile it yourself. No login. No telemetry. No monthly toll on your own work.
What it does
A full editor with an agent inside.
Edit by conversation
Tell it what you want ("cut the dead air, punch in on the reaction, add captions") and watch the timeline change. Multitrack, undoable, and always yours to fine-tune by hand.
Kinetic captions
Word-level transcription drives TikTok-style animated captions that land exactly on the beat of your speech. Style them once, reuse everywhere.
Generate footage with ComfyUI
Point it at your ComfyUI box and the editor plans shots, writes prompts, generates video, and assembles cinematic multi-shot timelines, including voice cloning and TTS workflows.
Motion graphics that write themselves
Twenty-plus built-in Remotion components. When none fit, the AI authors a new animated React component on the spot, smoke-tests it, and drops it on your timeline.
Lock in your brand
Styleguides pin your colors, fonts, and tone so every title card, infographic, and caption stays on-brand across every project.
Any model, including free local ones
OpenRouter, Anthropic, OpenAI, DeepSeek, GLM, or run LM Studio, llama.cpp, or vLLM locally and pay no one. Your keys, your machine, your footage.
Zero to posted
Four steps, no manual.
-
Connect a model
A guided first-run wizard sets up any of eight providers in about a minute, with presets, a live model list, and a connection test.
-
Pick a starting point
Record-and-caption, ComfyUI cinematics, brand styleguide, or infographic video. The assistant walks you through your first project.
-
Direct, don't drag
Chat to rough it in; use the full timeline editor to perfect it. Record webcam reactions, hunt clips from the web, stitch, zoom, transition.
-
Export and post
Bundled FFmpeg and a bundled renderer mean exports just work. Nothing else to install, ever.
Pricing
Two honest ways to get it.
The app
$79 one time
Yours forever. Free updates. No subscription, no account.
- Signed & notarized Mac app: download, drag, done
- FFmpeg, render engine, and caption pipeline bundled
- Guided setup wizard for eight AI providers
- All future 1.x updates included
- Support the roadmap with a single purchase
The source
$0 forever
The exact same editor, compiled by you, for personal & noncommercial use (PolyForm Noncommercial).
- Full source on GitHub, nothing held back
- Build with two commands (Bun + FFmpeg required)
- Hack it, extend it, learn from it
- Community support via issues
Questions
The fine print, minus the fine print.
Do I need a subscription?
No. Pay once and the app is yours: run it forever, offline, on your Mac. Bring your own AI provider keys (or free local models), so there's no metered middleman between you and your editor.
What's the difference between the free and paid versions?
Same code. The paid download is the convenience and the commercial license: a signed, notarized, one-click Mac app with everything bundled. The free path is the full source under PolyForm Noncommercial: build it and use it freely for personal and noncommercial work. Making money with it? That's what the one-time purchase covers.
Does my footage leave my machine?
Your media stays local. The only things that leave are the text (and, when you ask for vision features, frames) sent to the AI provider you configured, which can itself be a local model, making the whole pipeline private end to end.
Which AI models work?
Anything OpenAI-compatible plus native Anthropic support: OpenRouter, Anthropic, OpenAI, DeepSeek, GLM/Z.ai, LM Studio, llama.cpp, vLLM, Ollama. The setup wizard has presets and directions for each.
Do I need ComfyUI?
Only for AI video/image/audio generation. Everything else (recording, captions, motion graphics, editing, export) works without it.
What are the system requirements?
An Apple Silicon Mac on macOS 14 or later. FFmpeg, the render engine, and the caption pipeline all ship inside the app.