GPU-Accelerated Ingest
A CLIP vision model analyzes every clip on your own machine - on your NVIDIA GPU (CUDA) when available, or on the CPU. The resulting embeddings enable semantic search over your entire library.
GPU-Accelerated Ingest
A CLIP vision model analyzes every clip on your own machine - on your NVIDIA GPU (CUDA) when available, or on the CPU. The resulting embeddings enable semantic search over your entire library.
Beat-Synced Editing
A neural beat and downbeat model reads the beats, bar starts, energy curve, and drop zones out of your music. Every cut lands within ±200ms of a musical onset.
One Click to a Draft
Pick your videos and a song: New video analyzes them and cuts a draft of the whole song to the beat. Describe what you want in plain text, or render the finished video in the same go.
9 Built-In Presets
From RAW cuts to full PRESTIGE cinematography. Each preset controls pacing, transitions, VFX, and color grading.
Text Descriptions
Two text fields auto-generate a 4-tier semantic driver. No JSON knowledge required - power users can still use custom drivers.
100% Local
No cloud uploads. No telemetry. No subscriptions. One-time purchase. Your footage stays on your machine.