Compare
Head to head
Aria vs UBTech U1
Aria, from Realbotix, lists at F-Series from $125k · M-Series from $95k · bust from $20k. Launched at $175k.; UBTech U1, from UBTech Robotics, at U1 Lite (semi-torso) from ¥119,800 (~$17,650); full-body U1 Pro ¥169,800 (~$25,000); high-dynamic U1 Ultra up to ¥990,000 (~$146,500). Adults only.. Here's how the two compare across price, realism, specifications and availability.
| Status | Shipping | Pre-order |
|---|---|---|
| Type | Companion | Companion |
| Form | Full body | Full body |
| Presents as | Female-presenting | Male- or female-presenting |
| Price | F-Series from $125k · M-Series from $95k · bust from $20k. Launched at $175k. | U1 Lite (semi-torso) from ¥119,800 (~$17,650); full-body U1 Pro ¥169,800 (~$25,000); high-dynamic U1 Ultra up to ¥990,000 (~$146,500). Adults only. |
| Subscription | $199.99/mo | — |
| Availability | Shipping now | Preorder · ships Sept 2026 |
| Origin | USA | China |
| Realism index | 64/100 | 60/100 |
| Height | 175 cm · 5 ft 9 in (third-party) | Male 183 cm (6 ft 0 in) · Female 168 cm (5 ft 6 in) |
| Weight | ~54 kg (third-party) | Male 42 kg · Female 35.2 kg |
| Skin & materials | Platinum-cure silicone over an articulated frame | Proprietary biomimetic silicone skin; realistic hair; expressive face |
| Actuation | Electric servos — 17 facial motors, up to 44 DOF | 88 degrees of freedom; dual-pivot biomimetic neck; claims ~90% of basic human motion |
| Power | Built-in battery | Onboard rechargeable battery |
| Runtime | 4–8 hours (F-Series) | ~2–4 hours per charge |
| AI & conversation | Proprietary deep-learning conversation; AI-agnostic (cloud GPT or on-device) | Emotion-aware 'fast-and-slow' brain (100B+ params); ~500 ms reflexive response; 'Agent Memory OS' |
| Voice | Customisable voices | Conversational; speech-to-lip sync within ~20 ms |
| Mobility | Wheeled base — does not walk | Bipedal — walks, sits, stands on flat indoor floors; no stairs or rough terrain |
| Expressions | 17 facial motors; emotionally aware, tone-matching | Reads 20+ emotions from face, tone and speech (>90% claimed); adaptive expression |
| Languages | 100+ languages | Mandarin at launch; others undisclosed |
| Connectivity | Wi-Fi, Bluetooth | Wi-Fi; head-mounted cameras; encrypted memory stored on-device (Rockchip RK3588) |