What is Kinetix?
Driven by the mission to “pioneer cutting-edge 3D and human motion intelligence,” Kinetix combines machine learning with three-dimensional technology to build industry-leading generative models and core animation technologies. Its research focuses on human pose estimation, computer vision, and the foundational architecture of 3D character workflows, with a team composed of engineers, scientists, and strategic experts.
Key features
Kamo-1 is a 3D conditional video-generation model launched by Kinetix, supporting multimodal conditional inputs, including 3D camera trajectories, 3D character animations reconstructed from RGB performance videos, reference images, and natural language prompts, enabling coordinated control over both camera motion and character actions.
Kinetix’s video-to-animation model has been natively integrated into the Unity Muse AI service, empowering game developers to directly generate 3D animations; it has also been packaged as the ‘Emote Creator’ feature within the OVERDARE game, allowing players to create custom expression animations, accompanied by a content moderation system.
Use cases and audience
Game developers, real-time 3D content creators, game platform players, professional animators
- Game developers generate 3D character animations within Unity Muse
- OVERDARE players create custom expression animations using the built-in Emote Creator
- Content creators use Adobe Mixamo’s AI models to automatically rig characters and generate animations
RGB motion videos, 3D camera trajectories, 3D character animations, reference images, natural language prompts
AI-generated videos with physical consistency, featuring coordinated control of camera motion and character actions
Frequently asked questions
What input modalities does the Kamo-1 model support?
Kamo-1 supports multimodal input, including 3D camera trajectories, 3D character animations reconstructed from RGB video, reference images, and natural language prompts.
Has Kinetix technology been deployed in real-world commercial applications?
Yes, it is commercially available: integrated into Unity Muse for game developers; offered as the in-game Emote Creator within OVERDARE for players; and licensed for integration into Adobe Mixamo for content creators.
How to use
- 01
Input RGB motion video into the Kinetix video-to-animation model
- 02
Select a 3D character mesh and the target range of motions
- 03
Apply physical constraints and spatial layout control signals (such as depth and pose estimation)
- 04
Generate the final video with camera trajectories and coordinated control of character movements
Official website & sources
Sources
Updated on 2026-10-29
