Originally posted for the dev community interested in AI, robotics, and emerging tech
Hey devs! 👋
So something wild just happened in the robotics world that I think we should talk about. 1X Technologies (based in Palo Alto) just announced NEO—the first consumer-ready humanoid robot—and it's available for pre-order right now.
And yeah, I know. We've heard "the future of robotics is here!" before. But this one's actually different. Here's why.
What Is NEO? The TL;DR
NEO is a 66-pound humanoid robot designed to handle household chores. You can:
- Fold laundry
- Tidy up
- Water plants
- Fetch items
- Basically any household task you'd rather not do
Price: $20,000 to buy outright OR $499/month subscription
Shipping: 2026
Pre-order: $200 refundable deposit
But what makes it interesting to us as developers? The tech stack is genuinely impressive.
The Engineering That Actually Matters
Here's where it gets nerdy (in a good way).
Tendon Drive System
Instead of traditional harmonic drives (which are heavy and rigid), NEO uses 1X's patented tendon-drive system inspired by biological muscles. This isn't just aesthetic—it's functional:
- Highest-torque density motors in existence powering the tendon-based transmissions
- 22dB noise level (quieter than a modern refrigerator)
- Human-level dexterity with 22 degrees of freedom in each hand
- Incredibly safe because the architecture is inherently compliant
The whole system is covered in a custom 3D lattice polymer "soft suit" made from machine-washable nylon. This is the kind of thoughtful hardware design that usually only shows up in academic papers.
The AI Layer: Redwood Model
NEO runs on Redwood AI—1X's generalist AI model. Here's the important part: it doesn't ship fully autonomous.
Instead, it uses what they're calling "human-in-the-loop" training. Early adopters become part of the AI development process:
- You try to get NEO to do a task
- If it can't figure it out, you can enable "Expert Mode"
- A 1X operator pilots NEO remotely while you watch
- That session becomes training data for Redwood
- Over time, NEO learns from thousands of real-world interactions
This is actually brilliant. It's basically the same strategy Waymo uses for autonomous vehicles—supervised autonomy that gradually becomes full autonomy.
Why This Matters for Developers
1. Open Data for AI Training
Early adopters are essentially contributing to a massive AI training dataset. If you're interested in robotics + machine learning, this is the ground floor.
2. API Opportunities Coming
1X has announced that developers can apply for API access to build custom plugins for Windows apps and potentially add new capabilities to NEO. This could get wild.
3. It's Proof of Concept
Whether NEO succeeds or fails commercially, it proves humanoid robots can be consumer products. That's a category shift. Now other companies will follow.
The Technical Specs That Matter
Height: 5'6"
Weight: 66 lbs (29.94 kg)
Lifting capacity: 154 lbs (69.85 kg)
Carrying capacity: 55 lbs (24.95 kg)
Noise level: 22dB
Hand DOF: 22 (degrees of freedom)
Battery life: ~2 hours per charge
Self-charging: Yes (plugs itself in)
Communication: WiFi, Bluetooth, 5G
The self-charging part is actually important. It means you don't have to babysit it—when the battery gets low, it finds a charging dock and plugs itself in.
The Honest Reality Check
What's NOT Included:
- Full autonomy (it'll learn over time, but starts with basic autonomy for simple tasks)
- Superior resolution cameras (compared to Vision Pro, but that's not really the point)
- Real-time gesture control (it listens to voice, but no hand gestures yet)
- Unlimited battery (2 hours is your window)
What's Real:
- The engineering is genuinely innovative
- The transparency about early-adopter training is refreshing
- The price point ($20k) is actually reasonable for what you're getting
- The business model (buy or subscribe) gives flexibility
The Privacy/Security Question
Okay, this is important. NEO has cameras and mics. It watches your home. 1X hasn't been super detailed about:
- Exactly what data gets sent to their servers
- How encrypted the connection is
- What happens if the device gets hacked
- Data retention policies
They say they operate on a "big sister principle" (helpful) not "big brother" (invasive), but we've heard that before. If you're pre-ordering, this is worth asking about in detail.
Is This Worth $20,000?
For Most People: Probably not yet. Wait for generation 2.
For Early Adopters: Maybe. You're paying for:
- To be part of the AI training process
- The engineering achievement itself
- Potentially historic moment in robotics
For Accessibility: Actually yes. Someone with mobility issues or chronic pain? This could be life-changing.
What's Next?
1X is being cagey about the details, but they've hinted at:
- Improvements to autonomous capability over time (through data)
- API access for developers
- Different configurations (home vs. commercial use)
- Manufacturing scale-up (they're talking hundreds of thousands of units)
The Developer Angle
If you're interested in robotics, AI, or just watching how new technology categories get created, this is worth following. The code isn't open (yet), but:
- The engineering decisions are worth studying
- The human-in-the-loop training approach is industry-leading
- The business model is innovative
- The partnerships (OpenAI, Samsung NEXT, etc.) suggest serious backing
The Bigger Picture
We're watching something important happen. Robotics is moving from "research" to "product." Whether NEO succeeds or fails, it validates the category.
This is where your kids will be like "wait, you used to fold your own laundry?" and it'll sound ridiculous.
Should You Pre-Order?
Yes if:
- You're excited about robotics and willing to accept early-product issues
- You have the $200 to lock in a spot risk-free
- You want to be part of the training data generation
- You're genuinely interested in accessibility tech
No if:
- You expect it to work perfectly out of the box
- You're concerned about privacy in your home
- You want independent third-party reviews first
- You don't have the budget
Honestly? If you're reading this on dev.to, you probably fall into the "yes" category. We tend to be early adopters anyway.
Final Thoughts
NEO isn't perfect. It won't solve robotics overnight. But it's real, it's shipping in 2026, and it represents a genuine inflection point in how we interact with technology at home.
The next time someone asks "when will robots be mainstream?" you can point to this moment and say "October 2025."
Have you pre-ordered? Are you planning to? Drop your thoughts in the comments. I'm genuinely curious what the dev community thinks about this.
And if you're interested in the robotics/AI angle, I'd love to hear about projects you're working on!
Top comments (1)
The privacy concerns are real though. Who's watching my house through the robot's cameras? They need to be more transparent about this.