RADiCAL

The core of the tool is AI motion capture that works from ordinary footage. Film someone with a normal camera, upload the clip, and the system reconstructs the body's movement as 3D data you can apply to a character. No suit, no markers, no calibration rig. That removes the biggest cost and the biggest setup hurdle in traditional mocap, and it means a solo creator or a small team can capture motion anywhere they can shoot video. The tradeoff is accuracy: without markers, fine details like individual finger motion are harder to nail than on a dedicated stage.

Beyond a single character, the toolset can turn a video sequence into a 3D scene by rebuilding cameras, characters, and the surrounding space. This is useful when you want more than a skeleton: you get a scene you can move around inside, with the camera path and environment reconstructed too. For filmmakers doing previsualization, that means a rough but workable 3D version of a shot before the real one is staged. It is not a photoreal rebuild, so treat it as a planning and layout aid rather than a final environment.

Alongside motion data, the pipeline generates the pieces a compositor would normally build by hand: camera tracking, alpha masks, clean plates, and character passes. Camera tracking rebuilds the camera's movement in 3D so objects and effects sit correctly in the shot; alpha masks cut a subject out of the background; clean plates give you the background without the subject. Having these produced automatically saves hours of manual rotoscoping and tracking, and they come out editable so you can fix what the AI gets wrong.

The output is built to land in the software animators already use, not to be a closed box. Motion and scene data export to Maya, Blender, Unreal Engine, and 3ds Max through the USD format, a common exchange standard for 3D pipelines. That matters because it means the tool slots into a real production workflow: you capture and generate in the cloud, then finish in the app you know. Full creative control stays with the artist, and you are not locked into editing inside the browser.

Pricing runs on monthly credits rather than flat seat fees, so you spend on the project types you actually use. A free tier gives newcomers a set number of credits each month to explore AI motion capture, with caps on export resolution, storage, upload size, and the number of people captured at once. Paid tiers raise those limits and add more credits and higher-person capture. The free tier is a real way to test the tool, which matters for a category where you want to see the output quality before committing.

A common complaint about AI tools is that they act like a black box: one click, and you take what you get. This toolset is built so you can guide and refine the result instead, adjusting the output frame by frame rather than accepting a single take. In practice that means when the motion looks off, you fix it in your 3D app rather than starting over. The approach treats AI as an enabler for the artist, and the exported, editable elements are what make that possible.