Train and deploy your own open-source humanoid in Python
Hi HN, I’m Ben, founder of K-Scale Labs (YC W24). Last year, I wanted to buy a humanoid robot that I could hack on, but the few options for sale were either too expensive, proprietary, or had a limited SDK. We set out to build an affordable humanoid robot using off-the-shelf components that can be built and shipped today, capable of running modern machine learning models, and make it completely open-source for developers like me. Today, we’re releasing our reinforcement learning library and sim2real pipeline for people who want to train policies for humanoid robots. If you have a computer,…
In plain words
K-Scale Labs provides an open-source platform for training and deploying humanoid robots using Python. The system includes a reinforcement learning library and sim2real pipeline that allows developers to train robot policies on their computers before transferring them to physical hardware built from off-the-shelf components. Designed for developers who want affordable, hackable humanoid robots with modern machine learning capabilities, the platform can be tested in under five minutes and is entirely open-source.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hi HN, I’m Ben, founder of K-Scale Labs (YC W24). Last year, I wanted to buy a humanoid robot that I could hack on, but the few options for sale were either too expensive, proprietary, or had a limited SDK. We set out to build an affordable humanoid robot using off-the-shelf components that can be built and shipped today, capable of running modern machine learning models, and make it completely open-source for developers like me. Today, we’re releasing our reinforcement learning library and sim2real pipeline for people who want to train policies for humanoid robots. If you have a computer, you can try out this pipeline in less than 5 minutes: https://github.com/kscalelabs/ksim-gym. Or try on Colab: https://colab.research.google.com/github/kscalelabs/ksim-gym... Getting started is as easy as: git clone https://github.com/kscalelabs/ksim-gym.git cd ksim-gym pip install -r requirements.txt python -m train After training a model, you can send it to us and we will run it on one of our robots. We are building a benchmark for humanoid RL policies here: https://kscale.dev/benchmarks — Why does the world need another humanoid robot company? In the last year, humanoids have gone from science fiction to a seeming inevitability, bringing huge investment in the hardware supply chain and machine learning methods for robotics. But watching the ecosystem unfold, I felt pretty pessimistic about where things were headed. Seeing lots of cool demos without something that I can actually buy, from companies that have raised huge sums of money, reminded me of the early days of self-driving cars. On top of that, I find the idea of a small handful of companies building humanoid robots to be pretty dystopian. Even today, we’re seeing consumer robots being sold with government-mandated backdoors. That is not the future that I want to live in. To that end, our company has three long-term goals: 1. Ensure that the world’s best humanoid robots are white-box systems that anyone can program and audit 2. Create the infrastructure to radically simplify developer adoption of humanoid robots, paralleling CUDA for GPU programming or PyTorch for machine learning 3. Build an ecosystem to accelerate humanity’s transition to a post-scarcity Type 1 Kardashev civilization whose gains are maximally distributed If you would like to support us, you can pre-order one of our robots. We plan to launch the first robots this summer and are heavily discounting the price for early customers who can help us safely iterate on deploying robots in the wild: https://shop.kscale.dev Since we are focusing on creating a developer ecosystem, we would love to hear your thoughts and feedback about our current software and hardware stack: - Is this exciting to you? What would make you want to start developing on a humanoid robot? - What form factor is the most interesting for you (in terms of height, reach, end effector, or other hardware considerations)? - As a customer, what software capabilities would you expect from a humanoid robot in order to buy it? We would love your feedback!
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, May 2025
the whole month →
- C9
Life & fun · 2025 · felixrieseberg.github.io
