Clone Robotics · Humanoid · 2024
Clone Alpha, from Clone Robotics (Warsaw, Poland and the US Bay Area), is a musculoskeletal home android that takes a radically different approach from motor-and-gear humanoids — it mechanically replicates human anatomy. It's built on a polymer skeleton of 206 synthetic bones with articulated joints, moved by around 1,000 of Clone's proprietary 'Myofiber' artificial muscles that contract using water pressure (a single 3-gram fiber generates about 1 kg of force and contracts 30% of its length in under 50 ms), fed by a hydraulic 'vascular system' driven by a compact 500-watt electric pump likened to a heart. Its 'nervous system' comprises 4 depth cameras, 70 inertial sensors for joint proprioception and 320 pressure sensors for muscle-force feedback, with an NVIDIA Jetson Thor running Clone's Cybernet visuomotor model. Clone advertises a limited first batch of 279 units. IMPORTANT REALITY-FLAG: as of 2026 nothing has shipped, no price is confirmed, and the promised home chores are aspirations — the related Protoclone prototype has been demonstrated only suspended from wires and cannot yet stand or walk unaided. DOF (164) and component counts are company claims. Recorded as verified-manufacturer-claim with development status; no capabilities asserted as demonstrated that aren't.
Price on application
View full interactive profile, comparisons & videos →| Category | Humanoid |
| Sub-type | Musculoskeletal Android |
| Status | Development |
| Year | 2024 |
| Origin | Poland |
| Announced | 2024-12 |
| Released | 2025 |
| Weight | 35 kg |
| Frame | 206 polymer bones |
| Degrees of freedom | 164 |
| Battery life | 120 min |
| Camera | 4 depth cameras |
| Chipset | NVIDIA Jetson Thor |
| AI model | Clone Cybernet visuomotor model (end-to-end) |
| Operating system | Cybernet |
| Actuator type | Myofiber water-powered artificial muscles |
| Connectivity | Wi-Fi, Bluetooth |
| Use cases | Home assistance (planned), Biomimetic robotics research, Musculoskeletal AI development |
| Made in | Poland |
