Docs / Getting started / Overview

Overview

What the Animatica plugin for 3ds Max is, what it does in your scene, and how this host differs from the others.

Animatica for 3ds Max is a dockable tool window that brings motion generation into your Max scene. You describe the motion you want in plain language, and the plugin sends that request to an MMCP (Motion MCP) server — a local one on your machine, or Animatica Cloud — and keys the returned animation onto a skeleton in your scene.

The Animatica tool window docked to the left of the 3ds Max viewport, showing the Model and Skeleton cards with the Create Skeleton button, a freshly created bone skeleton in the viewport, and the Animatica Timeline running along the bottom of the screen.

#What it does in your scene

Builds or adopts a rig. Create Skeleton builds the canonical soma77 layout — 77 joints, rooted at Hips — under a namespace of your choosing, or you can point the plugin at a rig that is already in the scene.
Generates from text. Prompt blocks are laid out in time on the Animatica Timeline, and one generation turns all of them into a single request.
Steers the result with constraints. Six types: a full-body pose shape, four hand and foot effector pins, and a Path of ground waypoints.
Generates a single pose at the current frame, from one short prompt.
Captures motion from video — an .mp4 or .mov clip goes in, motion lands on the rig.
Loads motion from a file — .npz, .bvh or .glb.

Export is Max's own: File ▸ Export, and pick FBX. The plugin has no export path of its own, but it does ride on Max's export callbacks so a rig it built reads correctly in other applications.

#Three things worth knowing

The server does the generating. The plugin never launches a server and never generates locally; without a reachable server nothing can be produced.
The server also decides retargeting. Your rig goes on the wire as it is, and the server's own gate picks the fast path or its trained retarget model. If a server carries no retarget model and your rig does not match the model's canonical skeleton, the plugin refuses in words rather than producing something wrong.
Max is Z-up; the protocol is Y-up in meters. The conversion is done for you in one module, and your scene unit is read at runtime — you do not have to work in centimeters.

#How this host differs from the others

The plugin registers itself with the shared Animatica core under the host key max and declares exactly one capability: animation layers, which is real and detectable in Max. Capabilities default to absent, and the core gates its features on them rather than on a host's name — so anything that needs a capability this host does not declare is simply not in the window:

	3ds Max	MotionBuilder
Scene axis	Z-up, converted in one module	Y-up, no conversion needed
Scene units	asked at runtime, never assumed	—
Retargeting	decided by the server	decided by the server
Named animation containers	none — generation always replaces the scene's one animation	present
Character system / control rig	none — 3ds Max has no equivalent	present
Live drive	not implemented	present

Because Max holds one animation, there is no output-mode picker: the timeline header simply reads Mode: Replace. The one capability the host does declare is why Generate on base layer only exists in the Text to Motion card.

#Where to go next

Requirements and Installation — get the plugin running.
First launch — the Animatica menu, and pointing the plugin at a server.
The tool window — a map of the numbered workflow cards.

Verified against animatica-3dsmax-plugin@65c20ed