Head to head
Echo vs Grace
Echo (Realbotix, from $11,950) against Grace (Hanson Robotics, no public price), side by side on price, realism, specifications and availability.
The short version
- Echo scores higher on our Realism Index: 46/100 to Grace's 42. The widest gap is body & skin (58 vs 44).
- Echo has a public price ($11,950); Grace doesn't.
- Echo is on preorder; Grace is announced but not on sale.
- Echo also carries a $19.99/month software subscription.
Composed from the two directory records, so it updates when they do.
Echo46/100
Grace42/100
Realism Index sub-scores — six weighted dimensions built from public evidence. How we score
 | ||
|---|---|---|
| Status | Pre-order | Announced |
| Type | Companion | Humanoid |
| Form | Modular | Full body |
| Presents as | Female- or male-presenting | Female-presenting |
| Price | Echo Tabletop $11,950 · Echo Full Body $59,950 · $199 reservation deposit · AI plans from $19.99/mo · Vinci vision add-on $24,950 | Not a retail product; a 2022 report put the build cost near $35k per unit. Volume pricing undisclosed. |
| Subscription | $19.99/mo | — |
| Availability | Reserve now · ships H2 2027 | Prototype · care-home trials |
| Origin | USA | Hong Kong |
| Realism index | 46/100 | 42/100 |
| Height | Tabletop ~31-32 in (79-81 cm) · Full Body ~72 in / 183 cm (female), ~77 in / 196 cm (male) | — |
| Weight | Tabletop ~25 lb (11 kg) · Full Body ~88 lb (40 kg) | — |
| Skin & materials | Proprietary silicone-based synthetic skin on the face and neck (Tabletop); face, neck, forearms and hands (Full Body) | Frubber over a soft-skinned face and hands |
| Actuation | Tabletop 17 DOF · Full Body 39 DOF with articulated fingers; facial motor count not published | 48+ facial actuators; articulated arms; wheeled mobile base |
| Power | — | — |
| Runtime | — | — |
| AI & conversation | Realbotix AI plans; ChatGPT, Claude and Gemini integration claimed; your own models on Enterprise; on-robot speech and AI only on the top plan | Cloud AI for dialogue and preliminary screening; deep-learning emotion recognition |
| Voice | A distinct voice per character; chest-mounted speakers (stereo on Full Body) | English, Mandarin and Cantonese |
| Mobility | Stationary: the Tabletop sits on a desk; the Full Body stands on a base with upper-body movement only | Wheeled mobile base — does not walk |
| Expressions | "Advanced facial expressions" (maker's description); no motor count published | Expressive Frubber face; eye contact and gaze |
| Languages | 100+ languages; replies in the language it's addressed in | English, Mandarin, Cantonese |
| Connectivity | Realbotix App (subscription required); Full Body has dual cameras and four microphones | Wi-Fi; chest thermal camera for temperature/vitals; depth cameras |