CtrlK
| 1X NEO |
|---|
| General information |
| Manufacturer |
| Country of origin |
| Year announced |
| Type |
| Status |
| Predecessors |
| Variants |
| Price |
| Website |
1X NEO (stylized as NEO) is a bipedal humanoid robot developed by 1X Technologies, a Norwegian-American robotics company formerly known as Halodi Robotics. NEO was conceived as the company's first bipedal android platform, designed from the ground up for safe operation in private homes rather than industrial settings. The project was first announced publicly in March 2023 alongside a $23.5 million Series A2 funding round led by the OpenAI Startup Fund, with 1X stating that a physical demonstration of the robot would follow in summer 2023.[1][2] NEO made its first major public appearance at NVIDIA's GTC 2024 conference in March 2024, where a concept prototype stood on stage during Jensen Huang's keynote alongside humanoid robots from Boston Dynamics, Figure AI, Unitree, Agility Robotics, Apptronik, Fourier Intelligence, and Sanctuary AI.[3]
The NEO platform has progressed through several development stages: the original concept and early prototypes (2023 to early 2024), the NEO Beta engineering prototype unveiled in August 2024, and the consumer-oriented NEO Gamma redesign introduced in February 2025. On October 28, 2025, 1X opened consumer pre-orders for the production version of NEO at $20,000, with first deliveries to U.S. customers planned for 2026.[4]
NEO represents a strategic departure for 1X Technologies, which had previously built EVE, a wheeled humanoid robot deployed in commercial security and logistics. By transitioning to a bipedal, home-focused platform, 1X aimed to collect diverse real-world household data to train increasingly autonomous artificial intelligence models, following a philosophy that domestic environments provide a richer and more challenging training ground than factories or warehouses.[5]
1X Technologies was founded in May 2014 in Moss, Norway, under the name Halodi Robotics by roboticist Bernt Oivind Bornich, along with co-founders Phuong Nguyen, Jorgen Sundell, and Pal Loken.[6] Bornich, who studied Robotics and Nanoelectronics at the University of Oslo, had been interested in electromechanics since age 11 and envisioned building general-purpose robots that could work safely alongside people.[7] Loken, an investor, approached Bornich, Nguyen, and Sundell in 2014 with an interest in backing a robotics venture, and the four co-founded the company shortly afterward.[8]
In its early years, Halodi focused on developing fundamental hardware innovations for safe robotic actuation. The team rejected conventional high-ratio harmonic gears, which introduce friction, reflected inertia, and limited controllability. Instead, they developed the Revo1 motor, a proprietary high-torque brushless direct current (BLDC) motor paired with a cable-driven differential transmission. At the time of its introduction, the Revo1 was described as the highest torque-to-weight direct drive motor in the world.[8] The motor's low-friction, low-inertia design became the technological foundation for all of 1X's subsequent robots.
Development of EVE, Halodi's first humanoid robot, began alongside the company's founding in 2014. The first EVE prototype was publicly demonstrated in July 2017 performing simple warehouse and kitchen tasks.[8] EVE was formally introduced as a commercial product in 2018. Standing 188 cm (6 ft 2 in) tall and weighing 87 kg (192 lb), EVE was a wheeled, self-balancing humanoid robot designed for logistics, security, and healthcare environments. It used wheels rather than legs for locomotion and could travel at speeds up to 14.5 km/h (9 mph), carry loads of up to 15 kg, and operate for approximately six hours on a single charge. EVE featured 25 degrees of freedom and was powered by the company's Revo1 quasi-direct-drive actuators.[9]
EVE's most significant commercial deployment came through a partnership with ADT Commercial (later Everon). In 2022, 1X signed a contract to deliver 140 humanoid robots to ADT Commercial for autonomous night security patrols in commercial buildings across the United States, representing the industry's largest single order of humanoid robots for security applications at that time. Between 150 and 250 EVE units were ultimately deployed for night guarding duties. Third-generation EVE units were leased to ADT at approximately half a million Norwegian kroner per year.[10][11]
The operational data and lessons learned from EVE's real-world deployments, including insights about reliability, human-robot interaction, and the practical challenges of operating in unstructured environments, directly informed the development of NEO.[12]
In March 2023, Halodi Robotics officially rebranded as 1X Technologies, adopting a name that reflected the company's philosophy of building humanoids that work "at 1x speed" alongside people rather than replacing them.[13] The rebranding coincided with a strategic pivot from industrial and enterprise robotics to domestic consumer robotics. While EVE had demonstrated the viability of deploying humanoids in commercial settings, 1X's leadership concluded that the home market offered a larger long-term opportunity and, critically, a more diverse training ground for machine learning models.[5]
The NEO project was publicly announced in March 2023, when 1X revealed that proceeds from its $23.5 million Series A2 funding round would be used to build a new bipedal android model called NEO, alongside scaling manufacturing of EVE.[1] At the time of the announcement, 1X described NEO as a bipedal humanoid robot designed for domestic environments, featuring human-like motion and a slim anthropomorphic design intended to operate in the unstructured, unpredictable settings found in private homes.[2]
1X stated that a physical demonstration of the NEO robot would be revealed in summer 2023.[2] The concept represented a fundamental shift from EVE's wheeled design to fully bipedal locomotion, which would allow the robot to navigate multi-story homes, climb stairs, and move through spaces designed for human bipedal movement.
The original concept specifications described NEO as weighing approximately 25 kg (55 lb), which would make it considerably lighter than competing humanoid robots. For comparison, Tesla's Optimus weighed approximately 57 kg and Boston Dynamics' Atlas platform weighed around 89 kg at the time. This emphasis on extremely low mass was central to 1X's safety philosophy: a lighter robot carries less kinetic energy during any potential collision, reducing the risk of injury to people, pets, and household objects.[14]
From its earliest conception, NEO was designed around a set of principles that differentiated it from most humanoid robots under development at the time:
The design effort was led in part by Per Selvaag, who joined 1X as VP of Design bringing over 20 years of experience including roles as chief designer at Automobile Peugeot and lead designer at BMW. Selvaag's automotive and industrial design background shaped NEO's emphasis on human-centered aesthetics and psychological acceptability.[15]
The March 2023 Series A2 funding round was led by the OpenAI Startup Fund, with participation from Tiger Global, Sandwater, Alliance Ventures, and Skagerak Capital.[1] This investment was notable as one of the first major commitments by the OpenAI Startup Fund into a physical robotics company, signaling OpenAI's growing interest in embodied AI and the integration of large language models with real-world robotic systems.[16]
The relationship between 1X and OpenAI had begun in 2022, when the two organizations started collaborating on integrating large language models with physical robotic platforms. The investment provided 1X not only with capital but also with access to OpenAI's research in language understanding and task planning, which would later inform NEO's conversational capabilities and task decomposition systems.[17]
The OpenAI connection carried symbolic weight as well. At a time when Sam Altman and Elon Musk were engaged in an increasingly public rivalry, OpenAI's investment in 1X was interpreted by some commentators as a move onto Tesla's robotics turf, given Tesla's concurrent development of the Optimus humanoid.[18]
NEO made its first major public appearance at NVIDIA's GPU Technology Conference (GTC) in San Jose, California, in March 2024. During Jensen Huang's keynote address, a NEO concept prototype was displayed on stage alongside humanoid robots from several other leading robotics companies, including Boston Dynamics, Figure AI, Unitree, Agility Robotics, Apptronik, Fourier Intelligence, and Sanctuary AI.[3]
The GTC 2024 appearance was significant for several reasons. It placed NEO in the context of an emerging industry-wide push toward humanoid robotics, validated by NVIDIA's investment in robot foundation models and simulation tools. Huang used the demonstration to showcase NVIDIA's new foundation models for robots, which would enable humanoids to understand natural language and learn skills by observing human actions. While NEO at GTC 2024 was still a concept-stage prototype rather than a fully functional product, the public appearance generated substantial attention and positioned 1X alongside much larger and better-funded competitors.[19]
The conference also cemented a deepening relationship between 1X and NVIDIA that would later result in NEO adopting NVIDIA's Jetson Thor compute platform and a formal research collaboration announced in March 2025.[20]
In January 2024, 1X secured $100 million in Series B funding led by EQT Ventures, a European venture capital firm. The round included participation from Samsung Next, the OpenAI Startup Fund (following on from the Series A2), Tiger Global, and the Nistad Group.[21] The Series B brought 1X's total funding to approximately $137 million and valued the company at roughly $820 million post-money.[22]
The $100 million infusion dramatically accelerated NEO's development timeline. The funds were directed toward scaling the engineering team, expanding manufacturing capabilities at facilities in both Moss, Norway, and Hayward, California, and advancing the AI software stack that would power NEO's autonomous behaviors.
| Funding round | Date | Amount | Lead investor | Key participants | Valuation |
|---|---|---|---|---|---|
| Series A | 2021 | $10.1 million | Valinor | ADT Commercial, AutoStore founder Hatteland | Not disclosed |
| Series A2 | March 2023 | $23.5 million | OpenAI Startup Fund | Tiger Global, Sandwater, Alliance Ventures, Skagerak Capital | Not disclosed |
| Series B | January 2024 | $100 million | EQT Ventures | Samsung Next, OpenAI Startup Fund, Tiger Global, Nistad Group | ~$820 million |
| Reported raise | September 2025 | Up to $1 billion (in discussions) | Not confirmed | Not confirmed | $10 billion+ (reported target) |
By September 2025, reports emerged that 1X was in discussions to raise up to $1 billion in new funding at a valuation of at least $10 billion, representing more than a 12-fold increase from the Series B valuation.[23]
The first functional prototype of the NEO platform, called NEO Beta, was unveiled on August 30, 2024, with simultaneous events in San Francisco and Moss, Norway.[24] NEO Beta served as a proof-of-concept for 1X's tendon-driven actuation system and bipedal locomotion in home environments.
NEO Beta stood 165 cm (5 ft 5 in) tall and weighed 30 kg (66 lb), somewhat heavier than the original 25 kg concept target but still dramatically lighter than competing humanoids. It featured 55 degrees of freedom, hands with 22 degrees of freedom each, and could walk at up to 4 km/h (2.5 mph) with a sprint speed of 12 km/h (7.5 mph). The robot could carry payloads of up to 20 kg and operate for two to four hours on a single charge.[25]
The Beta's sensor suite included dual 8 MP fisheye RGB-D cameras with a 180-degree field of view, LiDAR, inertial measurement units, force-torque sensors, four microphones, and proximity sensors. The robot was clad in a soft jumpsuit with cushioned inserts positioned where human muscles would be, eliminating pinch points and making the robot safe for close physical interaction.[26]
A limited number of NEO Beta units were deployed internally at 1X offices and in selected employee homes for research and development purposes in late 2024. The Beta generated significant public attention, including a viral YouTube appearance in November 2024 showing a NEO Beta unit assisting in cooking a full steak dinner.[27]
| Specification | NEO Concept (2023) | NEO Beta (August 2024) | NEO Gamma (February 2025) | NEO Production (October 2025) |
|---|---|---|---|---|
| Height | ~165 cm | 165 cm | 165 cm | 167 cm |
| Weight | ~25 kg (target) | 30 kg | 30 kg | 30 kg |
| Degrees of freedom | Not disclosed | 55 | 75 | 75 |
| DOF per hand | Not disclosed | 22 | 22 | 22 |
| Walking speed | Not disclosed | 4 km/h | 5 km/h | 4 km/h |
| Sprint speed | Not disclosed | 12 km/h | 22.3 km/h | 12 km/h |
| Payload (carry) | Not disclosed | 20 kg | 25 kg | 25 kg |
| Battery life | Not disclosed | 2 to 4 hours | ~4 hours | Up to 5.5 hours |
| Operating noise | Not disclosed | ~32 dB | ~22 dB | 22 dB |
| Exterior | Soft covering (planned) | Soft jumpsuit with cushioned inserts | Seamless 3D-knitted nylon bodysuit | Seamless 3D-knitted nylon bodysuit |
| Compute | Not disclosed | Intel i7 + NVIDIA Xavier (inherited from EVE) | 1X NEO Cortex (NVIDIA Jetson Thor) | 1X NEO Cortex (NVIDIA Jetson Thor) |
| AI model | Not disclosed | Early prototype | Redwood (160M parameter VLA) | Redwood (160M parameter VLA) |
| Status | Concept | Engineering prototype | Consumer-ready redesign | Production model |
On February 21, 2025, 1X introduced NEO Gamma, a substantially improved iteration of the NEO platform designed to be consumer-ready.[28] NEO Gamma incorporated feedback from NEO Beta deployments and introduced improvements across virtually every subsystem:
NEO Gamma also introduced a new in-house large language model developed by 1X, replacing earlier reliance on external LLM providers for conversational interaction.[28]
On October 28, 2025, 1X officially launched NEO as a consumer product and opened pre-orders.[4] Two purchasing options were offered:
Customers could also place a fully refundable $200 deposit to reserve a unit. 1X announced plans to begin U.S. deliveries in 2026 and expand to international markets starting in 2027.[4]
The most distinctive engineering feature of the NEO platform, present from the original concept through all subsequent iterations, is 1X's proprietary Tendon Drive actuation system. This bio-inspired mechanism mimics the human musculoskeletal system by using artificial polymer tendons to transmit force from motors to joints, similar to how biological tendons connect muscles to bones.[29]
The Tendon Drive system evolved from the Revo1 motor technology developed for EVE. While EVE used cable-driven differential transmissions paired with low-friction motors, the NEO platform further refined this approach with higher torque density and greater compliance. The resulting system provides several advantages critical for a home robot:
Safety has been the central design constraint for NEO since the concept stage, given its intended deployment in homes alongside children, elderly individuals, and pets. Multiple layers of protection are integrated into the hardware:
| Safety feature | Description |
|---|---|
| Low mass | At 30 kg, NEO carries far less kinetic energy than heavier humanoids (57 to 89 kg range), reducing collision severity |
| Soft exterior | Custom 3D lattice polymer structures provide cushioning on all external surfaces |
| Knitted nylon suit | Machine-washable seamless bodysuit further reduces injury risk from contact |
| Pinch-free joints | All joints are fully enclosed, eliminating pinch hazards for fingers or skin |
| Compliant actuators | The Tendon Drive system's 95% backdrivability ensures joints yield under external force |
| IP protection | IP68-rated hands (fully submersible) and IP44-rated body (splash-proof) for kitchen and bathroom use |
The production NEO runs on the 1X NEO Cortex, a custom compute module based on NVIDIA's Jetson Thor system-on-chip. Jetson Thor uses the NVIDIA Blackwell GPU architecture and provides up to 2,070 FP4 TFLOPS of AI compute within a 130-watt power envelope. The chip includes 2,560 CUDA cores, 96 fifth-generation Tensor Cores, a 14-core Arm Neoverse-V3AE CPU, and 128 GB of LPDDR5X memory.[31]
NEO's intelligence is powered by Redwood, 1X's proprietary vision-language-action model. Redwood is a 160-million-parameter transformer that handles perception, navigation, and manipulation through a single unified end-to-end architecture. The model processes visual input from NEO's dual 8.85 MP stereo fisheye cameras and language instructions from users, then generates motor commands for the robot's entire body. Redwood runs entirely on NEO's onboard GPU at approximately 5 Hz without requiring a cloud connection, addressing both latency and privacy concerns.[32]
A key technical distinction of Redwood is its whole-body control approach: walking, arm movement, grasping, and spatial reasoning are all managed jointly by the same neural network rather than being handled by separate controllers. This enables coordinated behaviors such as bending at the hips and spine to pick up clothes from the floor, or bracing one hand against a surface for stability while opening a door with the other.[32]
1X also developed a separate World Model, built on a 14-billion-parameter generative video backbone. This model was trained first on web-scale video data, then adapted with 900 hours of egocentric human video, and finally fine-tuned on 70 hours of NEO-specific sensorimotor logs. The World Model enables NEO to learn new tasks from video data, even attempting actions the robot has never been explicitly trained on.[33]
To bridge the gap between current AI capabilities and fully autonomous operation, 1X employs a teleoperation-first strategy. When NEO encounters a task it cannot handle autonomously, a trained 1X operator wearing VR equipment can remotely access the robot's cameras and controls to complete the task. Each teleoperation session simultaneously generates training data that is fed back into the Redwood AI model, gradually expanding the range of tasks NEO can perform independently.[34]
Redwood is trained on a combination of real-world teleoperation data and autonomous rollouts. Human teleoperators remotely control NEO robots while the system records camera feeds and motor commands. The model learns from both successful and failed task attempts, allowing improvement from any interaction regardless of outcome. In addition, NEO was trained on 400 hours of "random play" data where the robot explored environments without specific task instructions, developing what 1X calls "common sense physics" for recovering from unexpected situations.[34]
1X CEO Bernt Bornich has described the approach as pragmatic: "If we don't have your data, we can't make the product better." The philosophy is that real homes present far more variety and unpredictability than laboratory environments, so collecting data from actual household deployments accelerates the path to full autonomy faster than any simulation-only approach could achieve.[34]
Privacy controls allow users to determine when and whether a teleoperator can access their robot's cameras and sensors. 1X offers scheduled sessions where an "Expert" teleoperator assists with complex tasks.[35]
In March 2025, 1X and NVIDIA announced a research collaboration focused on advancing humanoid robot autonomy. 1X created a dataset API for NVIDIA to access data collected from 1X offices and employee homes. The collaboration produced a demonstration for NVIDIA CEO Jensen Huang's GTC 2025 keynote, where NEO Gamma autonomously performed a dish-loading task using a policy built on NVIDIA's Isaac GR00T N1 foundation model. NEO Gamma also presented Huang with a custom studded leather jacket on stage during the event.[20][36]
NEO's adoption of NVIDIA's Jetson Thor as its compute platform further cements this partnership.
In December 2025, 1X announced a strategic partnership with EQT, one of the world's largest private equity firms (with approximately 267 billion euros in assets under management). The deal envisions the deployment of up to 10,000 NEO humanoid robots across EQT's portfolio of more than 300 companies between 2026 and 2030. Target use cases span manufacturing, warehousing, logistics, facility operations, and healthcare. Pilot deployments are expected to begin in the United States in 2026, with expansion across Europe and Asia to follow.[37]
This partnership is notable because it extends NEO beyond its original consumer home market into industrial applications, demonstrating the platform's versatility and providing 1X with a large-scale commercial revenue stream.
The relationship between 1X and OpenAI dates to 2022, when the two organizations began collaborating on integrating large language models with physical robotic systems. OpenAI's Startup Fund led 1X's Series A2 round in March 2023 and participated in the Series B in January 2024. While NEO initially relied on external LLM providers for conversational capability, 1X has progressively developed its own in-house language model to reduce dependence on third-party providers.[1][17]
1X manufactures NEO at facilities in Hayward, California, and Moss, Norway. In July 2025, the company consolidated its global headquarters into a new 80,000-square-foot facility in Palo Alto, California, capable of accommodating up to 400 employees. As of late 2025, 1X employs over 600 people.[38]
In August 2025, Vikram Kothari, who previously managed SpaceX's supply chain, joined 1X as Vice President of Operations to oversee the production scale-up.[38] The company's production roadmap envisions scaling from thousands of units in 2025 (for testing) to tens of thousands in 2026, with targets of 100,000 units by 2027 and millions by 2028.
NEO occupies a distinctive position in the humanoid robotics market as one of the few robots designed primarily for consumer home use. Most competing humanoids are significantly heavier and target industrial or enterprise applications first.
| Robot | Manufacturer | Height | Weight | Target market | Approximate price | Status (2026) |
|---|---|---|---|---|---|---|
| NEO | 1X Technologies | 167 cm | 30 kg | Home / consumer | $20,000 | Pre-order, shipping 2026 |
| Optimus | Tesla | 173 cm | 57 kg | Industrial / consumer | $20,000 to $30,000 (projected) | In development |
| Figure 03 | Figure AI | 170 cm | 70 kg | Industrial / enterprise | Not announced | In development |
| Atlas | Boston Dynamics | 150 cm | 89 kg | Research / industrial | Not for sale | Research platform |
| G1 | Unitree | 127 cm | 35 kg | Consumer / industrial | ~$16,000 | Available |
| GR-2 | Fourier Intelligence | 175 cm | 63 kg | Healthcare / industrial | ~$90,000+ | Limited availability |
NEO's key differentiators include its exceptionally low weight (30 kg versus 57 kg or more for most competitors), its bio-inspired tendon actuation for safe human interaction, its near-silent 22 dB operation, and its explicit consumer home focus. The $20,000 price point is among the lowest for a full-sized bipedal humanoid.[39]
NEO has received attention for its ambitious consumer home positioning and its lightweight, safety-first design approach. The high-profile backing from OpenAI and the major funding rounds have generated significant media coverage.
Critics and analysts have raised several concerns. Independent testing by journalists found that many household tasks required human assistance through 1X's teleoperation system rather than being completed fully autonomously, meaning early adopters are paying $20,000 for a robot with limited independent capability.[40] Privacy advocates have noted that the teleoperation model requires human operators to view live camera feeds from inside people's homes, even with user-controllable access settings.[41] The long-term viability of the $499/month subscription model has also been questioned given the robot's current limitations.
Supporters counter that the human-in-the-loop approach is a pragmatic strategy for collecting the training data needed to build truly autonomous home robots, and that early adopters are investing in a platform that will improve substantially over time through over-the-air software and AI updates.