What is the Menlo Asimov?
The Menlo AI Asimov is a general-purpose humanoid robot platform designed as a DIY kit for researchers and developers. Manufactured by Menlo AI in Singapore, it is named after the legendary sci-fi author, reflecting its openness to learning and exploration. The robot stands 120 cm tall, weighs 35 kg, and features 27 degrees of freedom with electric actuators and likely harmonic or planetary gears. Its onboard compute, likely an NVIDIA Jetson-class module, supports LLM integration, making it suitable for multimodal AI experiments. The frame uses aluminum alloy and composite parts, and it comes with 5-finger hands. As an open-source platform, it invites customization and modification for diverse research applications, filling a niche between costly commercial humanoids and small educational robots.
Specifications
Here are the full technical specifications.
| Spec | Value |
|---|---|
| Height | 120 cm |
| Weight | 35 kg |
| Degrees of freedom | 27 |
| Battery life | 1.5 hours |
| Max speed | 3 km/h |
| Payload | 35 kg |
| Price (new) | $15,000 |
| Price (used range) | Estimated $10,000–$14,000 (DIY kit, limited secondary market) |
Price & Value
New MSRP: $15,000
Used range: Estimated $10,000–$14,000
At $15,000, the Asimov undercuts most full-assembly humanoids and positions itself as a budget-friendly research entry. Its DIY format means significant additional investment in time and possibly extra components, but the open-source nature gives labs infinite flexibility. Resale value is contingent on the completeness of the kit; expect a 20–30% depreciation for well-maintained units. Compared to the $16,000 Unitree G1, the Asimov lacks factory integration and longer battery life, but its affordable price and hackability appeal to those willing to trade convenience for customizability. Total cost of ownership can rise if premium actuators or compute upgrades are added, but the baseline remains one of the cheapest options for a full-body humanoid. Buyers should budget for a compute module and sensors if not included in their kit package.
Who Is It For?
Best for: - University robotics labs and DIY enthusiasts wanting a customizable humanoid platform for embodied AI research - AI researchers exploring LLM-driven interactions and mobile manipulation experiments - Developers needing an open-source humanoid skeleton for software and control algorithm testing
Not for: - Industrial production lines requiring certified safety and continuous heavy-duty operation (1.5h battery, no formal certifications) - Enterprise deployments where turnkey operation and supplier support are mandatory - Users expecting plug-and-play operation — Asimov is a kit that demands assembly, programming, and troubleshooting
Alternatives & Comparison
The Asimov competes in the sub-$20,000 humanoid space, primarily with the Unitree G1 and the much more expensive Fourier GR-1. It differentiates through its open-source design and DIY kit approach, which appeals to budget-conscious researchers who prioritize modifiability over out-of-the-box polish.
| Model | Price | Available | Key Difference |
|---|---|---|---|
| Unitree G1 | $16,000 | yes | Fully assembled, 23 DOF, integrated sensors, 8-hour battery life — more plug-and-play but less open-source flexibility |
| Fourier GR-1 | $50,000 | enterprise-only | Higher payload (50 kg) and fully assembled industrial design, but significantly more expensive and not aimed at hobbyist/developer tinkering |
Verdict: For budget-conscious researchers and educators who need a blank canvas for humanoid experiments, the Asimov is unmatched in price and openness. Those who value immediate operation, longer runtime, and factory support should lean toward the Unitree G1. The Fourier GR-1 is overkill for Asimov’s target user, suited instead for well-funded labs with payload-heavy tasks.
Use Cases & Capabilities
Embodied AI Research Platform
With its onboard NVIDIA Jetson-class compute and LLM integration, the Asimov enables experiments in natural language interaction, vision-language-action models, and autonomous navigation. Researchers can deploy custom AI stacks to test long-horizon tasks like object fetching or multimodal scene understanding. The 27 DOF allow a wide range of arm and leg motions, making it suitable for locomotion and manipulation studies. Its open-source nature means algorithms can be tweaked at the actuator level, providing full access to control loops.
Robotics Education & DIY Building
As a DIY kit, the Asimov serves as a capstone project for advanced robotics courses, teaching students mechanical assembly, actuator calibration, and embedded programming. The 5-finger hands introduce grasping and dexterity concepts, while the modular frame encourages customization. Because it is built from standard components, learners gain insights into real humanoid design constraints like weight distribution and gear backlash. The community aspect of open-source hardware fosters collaborative troubleshooting and skill-sharing.
Lab Assistant for Light Tasks
The Asimov can perform simple fetch-and-carry tasks in a research lab, navigating between workstations at 3 km/h. Its 35 kg payload capacity is sufficient for moving lightweight equipment or sample trays. While its 1.5-hour battery limits continuous use, it can be used as an intermittent assistant for repetitive low-risk actions. Integration with LLMs allows it to respond to voice commands, aiding in routine lab management.
History & Background
Menlo AI, a Singapore-based robotics company, introduced the Asimov humanoid in 2024 as its first platform. Named after Isaac Asimov, the robot was conceived to democratize humanoid research by offering an affordable, open-source DIY kit. Prior to Asimov, Menlo AI focused on forward-looking robotics research; the company’s website highlights partnerships with organizations building robot workforces. Asimov debuted with 27 degrees of freedom, electric actuation, and onboard compute designed for LLM experiments. There have been no publicized generational updates as of 2026 — the current model remains the first generation. The robot has primarily been adopted by academic labs and hobbyist communities, operating on a direct-sales model from the manufacturer. Despite its limited public visibility, it has gained niche recognition in the open-hardware humanoid space.
Buying Used — What to Check
Kit completeness Missing actuators, compute board, or structural parts are common with DIY robots; verify all components listed in the original kit manifest.
Battery health The 1.5-hour runtime degrades over charge cycles; test for ability to hold charge under normal load.
Structural integrity Aluminum frame may show dents or stress fractures from experimental falls or modifications.