Animators should direct, not fill in frames.
A single second of hand-animated character motion can take an experienced animator hours. We build AI that generates it from a few key poses, keeping the animator in creative control.
Keyframes are sacred.
Every keyframe is a hard constraint. The model fills in around your intent. It never overrides it.
How our models are trained →Meet animators in their tools.
Native plugins, not another web editor. Generated motion lands on your armature, editable like hand-keyed frames.
Open at the core.
The MMCP protocol and generation backbone are open source: self-host it, extend it, build on it.
How the protocol works →Built in Poland.

Ex head of AI at TeaCode, previously a researcher at the University of Göttingen. Works on inverse problems and human motion modelling.

Ex head of CG at Platige Image; led cinematics for film and game franchises. Emmy nominee.
Animatica receives EU funding to advance AI for character animation, co-funded by the European Union.
