Explore by Humanoid Brand
TESLA
OPTIMUS
Tesla's general-purpose bipedal humanoid robot — built on the same AI and autonomy stack as Full Self-Driving. Designed for unsafe, repetitive, or boring tasks across factories, logistics, and eventually consumer homes.
In Production
"Gen 3" refers to upgraded hands only — body remains Gen 2 design (173 cm, 57 kg). * Deadlift target not officially confirmed.
FSD-Derived Intelligence
Optimus runs on an adaptation of Tesla's Full Self-Driving computer — trained on billions of vehicle miles. Eight cameras generating 576 megapixels per second feed end-to-end neural networks. No LiDAR, no radar. Gen 3 adds Grok for natural language task instruction and learning by video observation.
Fleet Learning
Every factory interaction trains the same central model. Task generalization improves at fleet scale — the same dynamic that accelerated FSD after broad deployment.
Dojo Supercomputer
Training runs on Tesla's Dojo cluster. The same infrastructure that trains FSD on road footage processes robot manipulation sequences.
Joint Torque Sensing
Every actuator includes real-time torque sensing. Foot sensors enable dynamic balance and slip detection. Gen 3 adds per-finger tactile feedback.
Learn by Observation
Gen 3 acquires skills by watching human demonstrations via video. Voice instructions via Grok allow task specification in plain English.
The Forearm Architecture
From 11 DOF to 22 DOF per hand — with all 25 actuators per side relocated into the forearm via a tendon-driven system, replicating human forearm muscle anatomy. Revealed February 17, 2026: "This bot got hands."
Where Optimus Works
Parts kitting, 4680 battery cell sorting, quality inspection. Currently deployed at Fremont and Giga Texas.
Warehouse picking, order fulfillment, goods transport. External commercial customers expected H2 2026.
Patient assistance, supply delivery. Long-term application; requires safety certification not yet in place.
House chores, errands, elderly care. Consumer units targeted 2028+ per Tesla's We, Robot framing.
DEPLOYMENT STATUS Q1 2026 — Optimus is not available for consumer or commercial purchase. Units are deployed internally for training data collection. External sales targeted H2 2026, consumer sales no earlier than late 2027.
Full development history → Tesla Optimus Development Timeline →
Tesla, Optimus, Tesla Bot are trademarks of Tesla, Inc. Androids.com is independent. Image credit: Courtesy of Tesla, Inc. Specs compiled from public filings, earnings calls, and verified press reports. Updated March 2026.
OPTIMUS
2021 – 2022
The origin story — from Elon Musk's first public announcement at AI Day 2021, through the unveiling of the first physical Gen 1 prototype at AI Day 2022. The year Tesla proved the concept was real hardware.
Part 1 Context — 2021 to 2022
When Musk announced Tesla Bot at AI Day 2021, the robot didn't exist — a human in a costume danced on stage as a placeholder. Twelve months later, physical hardware walked across a stage under its own power. This part covers those two foundational years: the concept announcement, the engineering sprint, and the first public proof that Tesla's humanoid program was real hardware, not vaporware.
Key dimensions: 5'8" tall (173 cm), 125 lbs (57 kg), max speed 5 mph. Specifically framed as unable to outrun a human — an intentional safety signal. The robot would use Autopilot cameras for vision and the FSD computer for inference. Musk said a prototype was expected "probably" in 2022.
Courtesy of Tesla, Inc.
The hardware used Tesla's FSD computer for inference and entirely custom-designed actuators — not off-the-shelf components. Walking speed was under 2 km/h, but the fact that any hardware existed less than 13 months after announcement established the seriousness of the program. The event triggered broad investor and competitor interest in humanoid robotics as a near-term commercial category.
What 2022 Established
By the end of 2022, Tesla had proven two things: it could build bipedal hardware fast, and it could leverage its existing AI and manufacturing infrastructure for robotics. The FSD computer, Autopilot vision stack, and in-house actuator manufacturing all transferred directly. The competitive moat — that no other car company had accidentally built a robotics AI company — was now physically demonstrated. Gen 1 was slow and limited, but it existed. The race was on.
OPTIMUS
2023 – 2024
The iteration years — from slow prototype to autonomous factory worker. Gen 2 debuts with 11-DOF hands, walking speed doubles, and by end of 2024 two units operate independently inside Tesla factories every single day.
Part 2 Context — 2023 to 2024
2023 and 2024 are the years that transformed Optimus from a slow-walking concept into a robot capable of autonomous factory work. Walking speed improved 4×, hand dexterity doubled twice, and by the end of 2024 two Optimus units were operating independently at Giga Texas daily. Gen 2 — revealed December 2023 — set the body design that still ships in 2026.
What 2023–2024 Established
By end of 2024, Optimus had transformed from a demo robot into an operational one. Two units worked autonomously in a real factory every day. The Gen 2 body — unchanged since December 2023 — was mature and stable. Walking improved 4×. Hands were heading toward 22 DOF. The first productive manufacturing task was on video. The question was no longer whether Optimus could work — it was how fast Tesla could make them.
OPTIMUS
2025 – 2026
The production era — from pilot line to Fremont mass manufacturing. Thousands of units targeted, Model S/X lines converted, Gen 3 hands revealed, and a firm consumer sales date of end-2027 publicly committed.
Part 3 Context — 2025 to 2026
This is the period where Optimus graduates from research project to manufactured product. The pilot production line goes live. Tesla commits $20 billion in CapEx. The Fremont Model S and X lines are converted. Gen 3 hands enter mass production. And Musk sets a public consumer sales date: end of 2027. The big caveat: Musk admitted on the Q4 2025 earnings call that current units are "primarily for learning, not productive tasks" — still in the R&D phase.
The substantive announcement: Tesla would end Model S and X production at Fremont to convert those lines for Optimus manufacturing — Tesla literally trading its two most iconic legacy vehicles for robot manufacturing capacity. Musk also hinted at crime deterrence and Neuralink consciousness upload integration use cases.
Fremont Model S and X production lines were being repurposed for Optimus manufacturing — Musk called it giving those vehicles an "honorable discharge."
The commitment: Tesla committed $20 billion in CapEx for 2026 — more than double the prior year. Targets confirmed: 10,000 units in 2025, 50,000 in 2026.
The candor: Musk acknowledged current units are "primarily for learning, not productive tasks" and confirmed the program is "still very much in the R&D phase." Notable honesty that tempered some of the mass production announcement momentum.
Results: lightweight hand (moves faster), better heat dissipation, per-finger tactile feedback.
Tesla simultaneously released a Weibo video: Gen 3 acquiring skills by observing humans — trainable via demonstration, voice, or video. No teleoperation. No pre-programming.
State of Optimus — March 2026
Mass production has commenced at Fremont with Gen 3 hands, $20 billion in CapEx committed, and a consumer sales date on the calendar. The honest picture: Musk's own Q4 earnings admission that units are "primarily for learning" means the operational threshold — robots doing genuinely productive work at scale — has not been crossed. The 2026 factory deployment results, and whether Gen 3 hands perform reliably in 24/7 conditions, will determine whether the 2027 consumer timeline holds.
Explore by Humanoid Brand
FIGURE
03
Figure's third-generation general-purpose humanoid robot — designed from the ground up for home environments and large-scale manufacturing. Purpose-built around the Helix vision-language-action AI and the first Figure platform engineered for high-volume production at BotQ.
Select Deployments
Vision-Language-Action Intelligence
Helix is Figure's proprietary vision-language-action (VLA) model — the AI system that transforms raw camera input directly into robot actions without intermediate programming. Rather than following scripted routines, Helix perceives the environment, reasons about it, and generates motor commands in real time.
Helix learns new tasks by watching humans perform them on video. Figure demonstrated towel-folding capability with only 80 hours of training footage — dramatically faster than traditional robotics development. Tasks demonstrated include: folding laundry, loading dishwashers, clearing tables, loading laundry, and navigating home environments while responding to voice commands.
Helix 02, released January 2026, extends Helix to full-body autonomy — coordinating all limbs simultaneously for complex whole-body manipulation tasks.
Pixel-to-Action Learning
Helix processes raw camera pixels directly into motor commands. No intermediate 3D reconstruction, no scripted motion libraries — the model learns the full perception-to-action pipeline end-to-end from demonstration data.
Fleet Learning via mmWave
Every Figure 03 unit uploads sensor data at 10 Gbps wirelessly during shift breaks. Insights from one robot's experiences improve the entire fleet — the more units deployed, the smarter all of them become.
Voice Command Interface
Helix supports natural language instructions. Operators can tell the robot what to do conversationally, and it reasons about how to execute the task given its current environment — no code, no configuration.
Helix 02 — Full-Body Autonomy
Released January 2026, Helix 02 extends the VLA model to coordinate all joints simultaneously for complex whole-body manipulation — enabling tasks that require coordinated arm, torso, and leg movement at once.
Built for Human Environments
Figure 03 is the first humanoid designed from the beginning with home safety as a primary constraint — not an afterthought. Every design choice reflects the reality of operating around people, children, and fragile objects in unpredictable domestic spaces.
Soft Textile Exterior
The entire outer surface is covered in soft, washable knit textiles padded with multi-density foam. Eliminates pinch points and significantly reduces injury risk compared to hard robot exoskeletons. Coverings are machine washable and customizable for commercial deployments.
Compact Form Factor
9% lower mass and smaller volume than Figure 02. Designed to navigate household tight spaces — hallways, doorways, between furniture — without the clearance requirements of industrial-scale robots.
3-Gram Tactile Sensing
Proprietary fingertip sensors detect forces as light as 3 grams — approximately the weight of a paperclip. Enables handling of fragile objects, delicate fabrics, and irregular-shaped items without breakage or damage.
Wireless Everything
No charging cables to trip over. No data cables to disconnect. Figure 03 steps onto a charging pad, connects to Wi-Fi, and handles power and data management autonomously. Designed for continuous unattended operation in a home setting.
Side Display Screens
New side-mounted screens allow quick robot identification across large commercial fleets. Fully customizable for branding, operational status, or fleet management data. A safety and logistics feature for enterprise deployments.
UL Safety Certification Targeted
The F.03 battery system is being developed toward UL-standard safety certification — the first Figure platform to explicitly target consumer electronics safety standards, a prerequisite for home deployment at scale.
What Figure 03 Does
AVAILABILITY Q1 2026 — Figure 03 is not available for direct consumer purchase. Select partner deployments are underway. Broader home availability targeted for late 2026. Early industrial deployments range from $30K–$150K based on Figure 02 precedent. Consumer target price ~$25,000.
2026
Full development history → Figure AI Development Timeline →
Figure, Figure 03, Figure AI, and all related trademarks are property of Figure AI, Inc. Androids.com is independent and not affiliated with Figure AI. Specs compiled from official announcements and verified press reports. Updated March 2026.
FIGURE AI
2023 – 2024
The breakout years — from emerging stealth to the world's first commercially deployed humanoid in a major factory. Figure 01 takes its first steps, the OpenAI partnership transforms the AI stack, and Figure 02 ships to BMW Spartanburg.
Part 2 Context — 2023 to 2024
2023 and 2024 are the years that took Figure from a stealth startup to the most talked-about humanoid robotics company in the world. Figure 01 emerged from stealth, took its first steps, won Intel Capital investment, and began commercial planning. Then 2024 delivered in rapid succession: an OpenAI partnership, a $675M raise backed by Jeff Bezos, Microsoft, and NVIDIA, the debut of Figure 02 with 16-DOF hands — and the first real-world commercial deployment at BMW. By end of 2024, Figure was shipping humanoid robots to a Fortune 500 paying customer.
The team of 40 engineers completed the full-scale humanoid build in six months before emerging publicly. The reveal included Figure's commercial vision: deploying robots in manufacturing, shipping, logistics, warehousing, and retail — industries with severe and worsening labor shortages. Figure 01 stood approximately 5'6" and was capable of lifting up to 44 pounds (20 kg).
All behaviors were learned, not teleoperated — a landmark distinction. The robot could insert a coffee capsule into a machine and press start after only 10 hours of watching demonstration videos. No explicit programming, no hand-coded sequences. This was the first public demonstration of a humanoid robot combining language reasoning with physical manipulation in a unified autonomous system.
Hands: New human-scale hands with 16 degrees of freedom and human-equivalent strength — enabling a far wider range of manufacturing and manipulation tasks.
Vision: Six RGB cameras with an onboard vision-language model for fast visual reasoning.
Compute: NVIDIA GPU-based modules offering 3× the AI inference power of Figure 01 — enabling fully autonomous task execution.
Battery: Custom 2.25 kWh pack — over 50% more energy than Figure 01, extending operational runtime significantly.
Figure 02 immediately began data collection and use-case training at BMW's Spartanburg facility.
The deployment at Spartanburg would eventually span 11 months, contribute to the production of 30,000 cars, and document 400% efficiency gains in specific tasks — one of the strongest real-world performance results ever recorded for a humanoid robot in a production environment.
Figure 02 · BMW Group Plant Spartanburg · Courtesy BMW Group
What 2023–2024 Established
By end of 2024, Figure had gone from emerging stealth to shipping humanoid robots to a Fortune 100 automotive manufacturer. The OpenAI collaboration produced the most compelling demonstration of humanoid AI reasoning seen publicly. The $675M raise from the biggest names in technology validated Figure's commercial direction. Figure 02 shipping to BMW proved that general-purpose humanoid robots could operate in real production environments — not just labs or staged demos. The stage was set for Figure 03 and Helix to take this to scale.
FIGURE AI
2025 – 2026
The scale era — Helix VLA transforms the AI stack, BotQ opens for mass production, Figure 03 launches as TIME's Best Invention of 2025, a $1B+ Series C values the company at $39B, and Figure 03 walks into the White House alongside Melania Trump.
Part 3 Context — 2025 to 2026
This is where Figure transitions from a well-funded startup shipping first-generation commercial robots into a company building toward mass deployment at scale. Helix — a vision-language-action model that learns from observation — becomes the central AI breakthrough. BotQ opens to manufacture robots at scale. Figure 03 launches as an entirely new platform purpose-built for Helix and home environments. A $1 billion-plus Series C at $39 billion valuation secures the financial runway. And on March 25, 2026, Figure 03 walks into the White House alongside the First Lady — becoming the most publicly visible humanoid robot in history.
Helix connects perception, reasoning, and motor control into a single end-to-end model — processing raw camera pixels directly into robot actions. Early demonstrations showed Helix accelerating real-world logistics tasks dramatically compared to previous programming approaches.
Instead of relying on contract manufacturers, Figure brought production of its most critical systems in-house. The facility features a proprietary Manufacturing Execution System (MES) with full traceability from subassembly through final assembly. BotQ is designed for Figure 03 — the first Figure robot engineered from the ground up for high-volume manufacturing.
This milestone was widely covered by CNET, The Robot Report, and major technology media as the clearest public demonstration of Helix's practical home-task capability.
Simultaneously: a strategic partnership with Brookfield for real estate and logistics deployments, and the announcement of Project Go-Big — Figure's initiative for internet-scale humanoid pretraining and direct human-to-robot transfer, aimed at dramatically accelerating the speed at which Helix learns new tasks by training on internet-scale video data.
Key advances over Figure 02: wireless inductive charging (feet on pad), 8 cameras including palm-mounted cameras, tactile fingertip sensors detecting 3-gram forces, soft washable textile exterior, 9% lighter and more compact body, 2× faster actuators, 10 Gbps mmWave data offload, and side display screens for fleet management.
Figure 03 was simultaneously named one of TIME Magazine's Best Inventions of 2025 — the first humanoid robot to receive the recognition.
A separate whistleblower lawsuit in November 2025 alleged safety risks during the BMW deployment period; Figure contested the claims. The core deployment results — 30,000 cars, 11 months, 400% efficiency — remained the headline commercial data point.
Figure 03 greeted attendees in 11 different languages, waved its hand, and introduced itself as "a humanoid built in the United States of America." Melania Trump said: "Figure 03, thank you for joining me. It's fair to state you are my first American-made humanoid guest in the White House."
The summit, focused on technology and children's education, was attended by first ladies of France, Sierra Leone, and Poland. Melania pitched a vision of robots as future educators. The footage went viral globally, accumulating hundreds of thousands of views within hours.
Figure AI had no active government contracts at the time. CEO Brett Adcock confirmed the appearance represented a significant endorsement for the company's US-built robotics mission.
State of Figure AI — March 2026
As of March 2026, Figure AI is the best-funded humanoid robotics company in the world at $39 billion valuation, with a dedicated manufacturing facility, a production-ready AI stack (Helix 02), and the most globally visible humanoid robot on the planet. The commercial track record — 30,000 BMW cars, 400% efficiency gains — is the strongest real-world performance data in the industry. Figure 03 is not yet available for consumer purchase, but select partner deployments are underway, and the company is targeting broader home availability for late 2026. The White House moment elevated public awareness of Figure and humanoid robotics globally in a single afternoon.
Explore by Humanoid Brand
UNITREE
HUMANOIDS
The world's most affordable humanoid robot lineup — R1 at $4,900, G1 at $13,500, H2 at $29,900, H1 at $90,000. Founded by Wang Xingxing in 2016, Unitree holds ~70% of global quadruped robot sales and shipped 5,500+ humanoid units in 2025. IPO targeting Q2 2026 at a $7B valuation.
The R1 is Unitree's entry-level humanoid — the most affordable full humanoid robot from any major manufacturer worldwide. Integrates multimodal large language models for voice and image understanding out of the box. Available in three variants: R1 Air ($4,900), R1 ($5,900), and R1 EDU (contact sales). Shipments began April 2026.
The G1 is Unitree's most popular humanoid platform — the best-selling humanoid robot from any manufacturer by volume. Compact at 127 cm, with extra-large joint movement range and optional 43-DOF configuration including force-control dexterous hands. Designed for imitation and reinforcement learning via NVIDIA Sim2Real. Performs backflips, dancing, martial arts, Tai Chi, and boxing in demonstrations.
The H2 is Unitree's full-size flagship humanoid for service environments. At 182 cm and 70 kg with a bionic dual-eye camera face, it is Unitree's most human-like robot. Features 31 DOF, 360 N·m max leg torque, and a 2070 TOPS onboard AI chip. Delivered its Spring Festival Gala 2026 debut as a "Sword Grandmaster" performing choreographed martial arts. Shipping from April 2026.
The H1 was China's first full-size general-purpose humanoid robot capable of running — it held the world speed record at 3.3 m/s and was the first full-size robot to perform a standing backflip. The H1 and the H1-2 (heavier, with expanded arm and ankle DOF) share an official product page. Target market: universities, research labs, state-owned enterprises. BYD and Geely have piloted H1 on production lines.
| Spec | R1 Air | G1 | H2 | H1 |
|---|---|---|---|---|
| Price | $4,900 | $13,500 | $29,900 | $90,000* |
| Height | 1230 mm | ~1270 mm | 1820 mm | ~1800 mm |
| Weight | ~25 kg | ~35 kg | ~70 kg | ~47 kg |
| Total DOF | 20 | 23–43 | 31 | varies |
| Max Speed | — | ~2.0 m/s | — | 3.3 m/s (record) |
| Max Leg Torque | — | — | 360 N·m | 360 N·m |
| AI Chip | Multimodal LLM | — | 2070 TOPS | — |
| Sensors | Vision + Voice | Vision + optional | Bionic dual-eye cams | 3D LiDAR + depth |
| Battery | — | ~2 hrs use | — | 864 Wh |
| Waist DOF | — | optional | 3 | — |
| Head DOF | 2 | — | 2 | — |
| Dexterous Hands | No | Optional EDU | Optional | Optional |
| Target Market | Consumer/Dev | Research/Edu | Service/Industry | Enterprise/Research |
| Shipping | Apr 2026 | Available | Apr 2026 | Available |
*H1 listed at $90,000 with "contact for real price." All specs from official Unitree pages. — = not publicly disclosed.
Wang Xingxing's Price-Breaking Machine
Unitree Robotics (Hangzhou Yushu Technology Co., Ltd.) was founded on August 26, 2016 by Wang Xingxing — a post-90s engineer who built his first quadruped robot, XDog, as a master's thesis project at Shanghai University in 2013. Wang left DJI to start Unitree in a 50 square meter office in Hangzhou's Binjiang District.
The company's defining competitive advantage is vertical integration: 90%+ of components — motors, reducers, encoders, sensors — are developed and manufactured in-house. This is why Unitree's prices are 30–50% below competitors. The company has been profitable every year since 2020, generated over ¥1 billion (~$140M) in annual revenue, and holds approximately 70% of global quadruped robot sales.
In 2025, Unitree raised a Series C round at approximately $1.7 billion valuation, backed by Alibaba, Tencent, China Mobile, Ant Group, Geely Capital, and HongShan Capital. In March 2026, Unitree received approval for an IPO on Shanghai's STAR Market to raise $610 million, targeting a $7 billion valuation.
NOTABLE: Wang Xingxing was the only entrepreneur of his generation invited to a February 2025 symposium of business leaders chaired by President Xi Jinping, seated in the front row near Huawei's Ren Zhengfei — a significant signal of Unitree's strategic importance to China's technology ambitions.
Sim2Real + NVIDIA + Open Source
Unitree trains its humanoids using NVIDIA's technology — digital twins learn movements through trial and error in simulation, then the behaviors are transferred to physical robots via Sim2Real. In 2026, Unitree released two open-source AI frameworks: UnifoLM-WMA-0 (a world-model-action architecture for predicting and adapting to future environmental states) and UnifoLM-VLA-0 (a vision-language-action framework combining visual perception, natural language understanding, and robotic action).
Sim2Real Training
Digital twins train in NVIDIA simulation environments, then behaviors transfer to physical hardware. Has produced viral demonstrations of G1 performing backflips, martial arts, Tai Chi, and boxing — all learned, not hand-programmed.
UnifoLM-VLA-0 (Jan 2026)
Open-source vision-language-action model released January 29, 2026. Combines visual perception, natural language understanding, and robotic action in a single framework — enabling developers to build applications on top of Unitree hardware.
UnifoLM-WMA-0 (Sept 2025)
Open-source world-model-action architecture that allows robots to predict future environmental states and adapt behavior accordingly — enabling more generalizable autonomous operation in real environments.
R1 Multimodal LLM
The R1 integrates multimodal large models for voice and image out of the box — dramatically lowering the development threshold for the cheapest humanoid on the market. Developers can build on R1 independently without robotics expertise.
Full development history → Unitree Development Timeline →
Unitree, G1, R1, H2, H1, H1-2, and all related trademarks are property of Hangzhou Yushu Technology Co., Ltd. Androids.com is independent. Specs from official Unitree pages and verified press reports. Updated March 2026.
Explore by Humanoid Brand
NEO
BY 1X
The world's first consumer-ready home humanoid robot — designed from the ground up for safe domestic use. NEO automates household chores, provides natural voice interaction, and learns continuously through real-world use and Expert Mode teleoperation. Shipping to US homes in 2026.
Pre-Orders Open
AI That Learns From Life
NEO runs on 1X's proprietary AI stack built around two core components: the 1X World Model — a learned simulator trained on over 1 million hours of video that predicts future states and enables generalization to new environments — and Redwood AI, a vision-language model that controls NEO in real time.
Unlike traditional robotics approaches that require explicit programming, the 1X World Model enables NEO to observe a new situation, reason about it, and attempt unfamiliar tasks even without prior training on that specific scenario. The AI continuously improves through OTA software updates, meaning the robot gets smarter the longer you own it.
1X is backed by the OpenAI Startup Fund — the only robotics investment OpenAI has ever made — giving the World Model conceptual DNA shared with the most advanced language AI systems in the world.
1X World Model
A learned simulator that predicts future environmental states based on sensor input. Enables NEO to reason about novel situations and generalize from training data to real-world scenarios it has never explicitly encountered before.
Redwood AI
The real-time vision-language model that translates World Model reasoning into physical motor commands. Processes camera input and audio simultaneously to control all joints in coordinated, smooth motion.
Expert Mode
When NEO encounters a task it cannot complete autonomously, a trained 1X operator takes remote control via VR headset, completes the task, and that interaction becomes training data — improving the AI for all future situations of that type.
Audio + Visual Intelligence
Built-in large language model, audio intelligence, visual intelligence, and persistent memory enable natural conversational interaction without relying on a screen. NEO understands context, remembers preferences, and responds accordingly.
Built for Real Homes
Chore Automation
Give NEO a list of chores, schedule when you want them done, and come home to a cleaner house. Demonstrated tasks include folding laundry, organizing shelves, tidying spaces, loading dishwashers, and fetching items.
Natural Voice Interface
Speak to NEO conversationally to access all features. The onboard LLM understands natural language requests, answers questions, and can carry on contextual conversations without waking words or app interactions.
Soft Body Safety Design
Custom 3D lattice polymer body, low-inertia tendon drives, and pinch-proof joints make NEO safe around people, pets, and children. Machine-washable soft good coverings keep the robot clean and refreshed. 22 dB operation — quieter than your refrigerator.
Self-Charging
NEO plugs itself in when it needs power — no human intervention required. 6 minutes of charge delivers one hour of runtime. The 842 Wh battery supports 4-hour operational sessions between charges.
Mobile App
Manage schedules, monitor NEO remotely, communicate in real time, and pilot the robot through unfamiliar situations via the companion app. VR remote control also available for detailed task guidance.
IP68 Hands
Fully dust and water protected hands (IP68) enable NEO to handle dishwashing, cleaning wet surfaces, and other tasks involving water. Body-level IP44 splash resistance rounds out home-safe protection.
Pre-Orders
Now Open
NEO is the only consumer humanoid robot with real shipping dates and transparent pricing. US and Canada deliveries begin Q3–Q4 2026. Europe and select Asian markets follow in 2027.
Pre-orders "far exceeded" 1X's internal goals per company spokesperson · EQT partnership commits up to 10,000 NEO units to global industry 2026–2030
Full development history → 1X Technologies Development Timeline →
1X, NEO, NEO Beta, NEO Gamma, Redwood AI, and all related trademarks are property of their respective owners. Androids.com is independent and not affiliated with 1X Technologies. Specs from official 1X pages and verified press reports. Updated March 2026.
Androids Origin/History
- Android (Greek: "man-like") = humanoid robot/automaton.
- First recorded 1720s ("androides") for mechanical humanoids.
- 1730s–1830s: dictionaries, patents, literature.
- 1886: sci-fi debut (Tomorrow's Eve).
- 1930s: pulp fiction (fleshy android vs. robot).