What is Kinetix?
Kinetix is a research laboratory focused on frontier 3D and Human Motion Intelligence. The company combines deep expertise in machine learning and 3D to develop state-of-the-art generative models and core technologies for character animation.
Core Technology
The flagship product is Kamo-1, a 3D-conditioned video generation model designed to bring physical grounding and precise control into AI video generation. It supports multimodal conditioning including 3D camera trajectories, 3D character animations reconstructed from RGB videos, reference images, and natural language prompts.
Instead of treating the scene as unconstrained pixels, Kamo-1 uses depth and pose estimates as control signals. This produces video that is physically consistent with the underlying spatial layout and articulated motion. The model enables coordinated control over both camera motion and character motion within the same generated shot.
Real-World Applications
Kinetix technologies are already powering major experiences:
- Native integration of the video-to-animation model into Unity Muse AI services for game developers.
- «Emote Creator» feature in KRAFTON’s OVERDARE game, allowing players to create custom animations with a moderation system.
- Licensing and integration into Adobe Mixamo to revamp character animation with AI models.
Research Foundation
The lab’s research focuses on human pose estimation, computer vision, and foundational components of 3D character workflows. Kinetix excels at both understanding and generating humanoid motion using 3D and machine learning.
Advantages and Limitations
Advantages: exceptional physical accuracy, precise multi-modal control, proven integrations with industry leaders (Unity, Adobe, KRAFTON), strong scientific foundation.
Limitations: primarily B2B and enterprise focused, no public consumer-facing tool or free tier is currently offered, high technical complexity.
Kinetix stands at the cutting edge of combining 3D intelligence, computer vision, and generative AI to create physically plausible character animation and video content.
