NEO

Home helper for cognitively-loaded chores

NEO is positioned as the first mass-market consumer humanoid. CEO Bernt Bornich has repeatedly said the wedge use case is laundry, dishwasher loading, tidying counters, and trash. Tendon-driven actuation makes NEO lighter (30 kg — half of Apptronik Apollo at similar height) and safer around people and pets: a runaway servo can't produce the impact force of a rigid actuator. Public demos through 2024-2025 cover pouring drinks, watering plants, folding laundry — slower-than-human but without safety enclosures. Manufacturer has not disclosed reliability rates; 1X markets teleoperator fallback for tasks the on-board policy cannot complete.

Consumer-scale data collection for the World Model

Bornich stated on the Robotics Business Review podcast Q4 2024 that customer NEO units are distributed data-collection nodes for 1X's proprietary World Model foundation model. Every household expands the training corpus for whole-body manipulation and home navigation in ways Skild AI's or 1X's own Redwood datasets cannot — data captured in real homes, not a lab. That is the economic bet behind the $499/month subscription: it lowers customer capex while preserving 1X's telemetry access over years.

Alpha program and early-access developer units

Before general availability, 1X ran a closed alpha program from Q3 2024 — small number of units with hand-picked customers (some ex-Tesla and ex-Waymo per Rest of World early 2025). Purpose: refine on-board policy, harden operations pipeline, generate case-study material. Alpha units run the same hardware but with heavier remote-supervision — each 1X operator intervention becomes training data.

Public demo circuit and brand-building

NEO's coming-out was a viral CES 2025 walking demo. Subsequent appearances: TED 2025 with Bornich on stage, tech-industry conferences, Norwegian government demos in Oslo. Unlike Tesla Optimus which stays under Tesla's PR control, 1X lets third parties film NEO in unscripted settings — hallways, patios, mixed lighting. Each unscripted clip reinforces the durability narrative and generates organic lift 1X does not pay for.

Not a research platform — no ROS, no open SDK

NEO ships with a proprietary stack: no ROS support, no exposed SDK for arbitrary policies, no third-party sensor integration path. This is the opposite of the Unitree G1 or H1 approach and rules NEO out for university robotics research and for foundation-model data collection by anyone other than 1X. Buyers who need programmability should look at G1 or H1-2 — NEO is a shipped consumer appliance, not a research platform.

NEO Beta unveil — August 2024

First bipedal from 1X after years of wheeled-base EVE deployments. 168 cm, 30 kg, 75 DOF, tendon-driven. 842 Wh battery, 240 min rated runtime, no hot-swap. Vision-only, no LiDAR. Public walking demo on Sunnyvale streets at unveil.

Alpha program launch — Q3 2024

Closed alpha with a small number of hand-picked customers in the Bay Area and Oslo. Units placed with tele-supervision agreement — 1X operators tele-assist when on-board policy fails. Intervention data feeds directly into World Model training.

Consumer pre-order open — October 2024

Pre-orders opened at $20,000 outright or $499/month subscription. Ship date not disclosed. By early 2025 the reported waitlist exceeded 5,000 units per Rest of World. Subscription preserves 1X's telemetry access.

CES 2025 walking demo — January 2025

Walking demo at CES 2025 drew the largest crowd of any humanoid on the floor per Robotics 24/7. Bornich keynoted the humanoid track and disclosed the World Model policy training runtime.

Ongoing through 2026 — Ongoing

GA date not disclosed. Bornich says 'measured rollout' — first commercial units expected in 2026, no firm month. Alpha program continues and reportedly expands into Europe via retail-chain showroom placements.

1X HQ — Sunnyvale, California and Oslo, Norway

Primary R&D facilities. Sunnyvale for policy training and demo builds. Oslo (1X's Norwegian roots) for hardware R&D and European operations. Internal fleet size not disclosed.

Source: https://www.1x.tech/

Alpha program customer homes — USA and Norway

Small number of hand-picked customer households running NEO with tele-supervision. Locations not publicly named. Data streams back to 1X for World Model training.

Source: https://restofworld.org/

Pre-order backlog

Reported 5,000+ pre-orders through early 2025 per Rest of World. 1X has not officially confirmed the number. Ship date remains disclosed as 'measured rollout' with no firm month.

Source: https://restofworld.org/

Trade shows and demos — CES, TED, and Norwegian government events

NEO appears at select high-profile events for brand building. CES 2025 walking demo drew largest humanoid crowd per Robotics 24/7. TED 2025 keynote with Bornich. Norwegian government demos in Oslo tied to 1X's national-industry positioning.

Source: https://www.robotics247.com/

Tendon-drive service life is unknown

Tendon-driven actuation is what makes NEO lightweight and safe — the trade-off is that tendons wear differently than rigid actuators. Manufacturer has not disclosed expected service intervals or replacement costs. Early alpha reports (unconfirmed) mention tendon retensioning as a routine service task. Buyers signing up for a $499/month subscription are protected; outright $20,000 buyers should factor in unclear long-tail service economics.

Vision-only, no LiDAR

NEO ships without LiDAR — 1X positions vision-based perception as sufficient for indoor consumer contexts. This works in well-lit, uncluttered spaces. In low light, high-contrast lighting, or cluttered environments (piled laundry, dark hallways) manufacturer has not disclosed reliability numbers. Users needing outdoor operation or unstructured environments should look at H1 or Digit — both ship with LiDAR by default.

1.4 m/s max speed is slow by design

Consumer safety envelope caps NEO at 1.4 m/s — slower than average human walking pace. In a home this is fine. In any emergency scenario (fire evacuation, dropped item rolling toward stairs), NEO cannot match human response time. Explicitly not designed for urgent tasks.

Proprietary stack, no ROS, no open SDK

Zero SDK for arbitrary policies. No third-party sensor add-ons. No compatibility with MuJoCo, Isaac Sim, or Gazebo tooling. Buyers who want to build custom applications should look elsewhere.

Ship date remains undisclosed

Pre-order opened October 2024. As of early 2026 no general customer has received a unit. 1X says 'measured rollout' — this is honest signaling, but buyers paying $20,000 today are effectively pre-funding a startup with no delivery guarantee. Subscription buyers face lower risk since they only pay upon delivery.

vs. 1X EVE

EVE is 1X's older, wheeled-base commercial platform — already shipping to enterprise customers including BMW facility trials since 2024. EVE is taller (186 cm), heavier (86 kg), and priced for B2B (not publicly listed but reported six-figure per Rest of World). NEO is bipedal, lighter, cheaper, consumer-oriented, and later to market. Different segments: EVE ships now for enterprise; NEO is the consumer bet. 1X operates them as complementary — data from EVE deployments feeds into NEO's World Model, and vice versa once NEO scales.

vs. Tesla Optimus Gen 2

Optimus Gen 2 targets a similar $20,000-30,000 consumer price point but has not opened public pre-orders as of early 2026. Tesla's vertical integration on actuators and battery could undercut NEO on unit economics, but the gap between announcement and shipping is unbounded. Buyers who need hardware in 2026 should look at NEO — pre-order is real, subscription available. Buyers who trust Tesla's execution and can wait have Optimus in the pipeline.

vs. Figure 02

Figure has committed to B2B — BMW manufacturing partnership since 2024, no consumer product on roadmap. Figure 02 at 70 kg is more than double NEO's mass and rated for industrial payloads. Different sub-segment: Figure is a factory robot; NEO is a home robot. Buyers evaluating between them are probably not the same buyers. NEO wins on safety envelope and consumer price; Figure 02 wins on industrial reach and payload.