Genesis AI releases Genesis World 1.0, a platform that evaluates robots 400 times faster
Genesis AI has released Genesis World 1.0, an AI-based platform for simulating and evaluating robots. The system cuts algorithm evaluation time from 200+ hours to half an hour and matches real robots at 90%. The company says this should speed up the development of scalable robotics.
AI-processed from MarkTechPost; edited by Hamidun News
Genesis AI has released Genesis World 1.0, a neural-network-based platform for simulating robots. The release came on May 27, 2026.
What the platform consists of
Genesis World is a four-component system for quickly evaluating robotics built on foundation models. It consists of a physics engine (realistic physics simulation), a renderer (visualization), a compiler (code transformation) and a set of tools for developers.
- A physics engine for accurate simulation of dynamics
- A renderer for realistic scene visualization
- A compiler for optimizing code and models
- Tools for researchers and developers
A 400x speedup
Genesis World's main achievement is cutting evaluation time from 200+ hours to under 0.5 hours. That is equivalent to a 400x speedup. Where a researcher once waited a week for test results on a single algorithm, the answer now arrives in minutes.
The system also reaches a correlation of 0.8996 (Pearson scale) between simulation and real-world robot performance. This means an algorithm that performs well in Genesis World is highly likely to work on a physical robot too.
What this means
Genesis World 1.0 lowers the barrier to developing scalable robotics. Until now, the main difficulty has been long testing cycles: even a simple algorithm change required hours of computation. Researchers can now test hypotheses and iterate quickly.
This matters especially for foundation models in robotics: large neural networks trained on heterogeneous data that can adapt to different tasks. Genesis World makes it possible to assess how such a model will behave on a specific robot without having the physical hardware.
Key Questions Answered
What the platform consists of?
Genesis World is a four-component system for quickly evaluating robotics built on foundation models. It consists of a physics engine (realistic physics simulation), a renderer (visualization), a compiler (code transformation) and a set of tools for developers. A physics engine for accurate simulation of dynamics A renderer for realistic scene visualization A compiler for optimizing code and models Tools for researchers and developers
What are the key details on A 400x speedup?
Genesis World's main achievement is cutting evaluation time from 200+ hours to under 0.5 hours. That is equivalent to a 400x speedup. Where a researcher once waited a week for test results on a single algorithm, the answer now arrives in minutes. The system also reaches a correlation of 0.8996 (Pearson scale) between simulation and real-world robot performance. This means an algorithm that performs well in Genesis World is highly likely to work on a physical robot too.
What this means?
Genesis World 1.0 lowers the barrier to developing scalable robotics. Until now, the main difficulty has been long testing cycles: even a simple algorithm change required hours of computation. Researchers can now test hypotheses and iterate quickly. This matters especially for foundation models in robotics: large neural networks trained on heterogeneous data that can adapt to different tasks. Genesis World makes it possible to assess how such a model will behave on a specific robot without having the physical hardware.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.