FusionMax
About
About FusionMax
FusionMax is Muks Robotics' two-billion-parameter audio-vision-language-action mixture-of-experts model for the Spaceo platforms. It combines RGB and depth perception, spatial audio, proprioception, language processing and motor output, activating specialised expert modules based on the input and robot state.
Muks describes task decomposition, long-horizon planning and generative action control, with deployment either on embedded NVIDIA hardware or a local server over a high-speed network. The model is proprietary and its current public evidence consists primarily of company descriptions and robot demonstrations. Claimed adaptability and human-level fluidity are not treated as independently benchmarked performance.
Capabilities
Source-backed capabilities
- Workflows: Perception, task decomposition, embodied reasoning and motor-control generation for humanoid robots.
- Operating environments: Muks humanoid development, service environments and controlled robotics deployments.
- Target markets: Organizations evaluating Muks Robotics humanoids or physical-AI software for robotic programs.
- Deployment model: Platform-integrated model layer for Muks robots, with on-device or local-server inference options.
- Natural-language interaction
Claimed - Breaks a multi-step spoken or textual instruction into subtasks and generates motor-control commands. - environment: Muks humanoid deployments
Source
Context
Market context
FusionMax is Muks Robotics' AVLA physical-AI model layer for Spaceo humanoids.
- Target environment: Muks humanoid development, service environments and controlled robotics deployments.
- Workflow context: Perception, task decomposition, embodied reasoning and motor-control generation for humanoid robots.
- Customer context: Organizations evaluating Muks Robotics humanoids or physical-AI software for robotic programs.
- Deployment model: Platform-integrated model layer for Muks robots, with on-device or local-server inference options.
- Commercial maturity: Commercially presented sensor architecture product with maturity tied to customer deployments, integration performance, and manufacturing support.
- Adoption constraints: Adoption depends on Spaceo platform maturity, embedded compute, latency and proof from real customer deployments.
- Market position: Humanoid model layer; closer to Helix and GENE-26.5 than to generic vehicle autonomy stacks.
- Adjacent products: Helix, GENE-26.5, sudo R1, FluxVLA Engine
Comparable Systems
- Being-H0.8
BeingBeyond - Robotics Foundation Model
- Helix
Figure AI - Robotics Foundation Model
- GEN-1
Generalist - Software
- GENE-26.5
Genesis AI - Robotics Foundation Model
- Lumo 2
Astribot - Robotics Foundation Model
- Rho-alpha
Microsoft - Robotics Foundation Model
Related Systems
- DYNA-1
Dyna Robotics - Robotics Foundation Model