Best AI Simulation tools
7 live tools in this category, ranked by real-world traffic.
Turn a single image or video into a persistent 3D world that can be explored endlessly, without the AI “forgetting” areas it has already visited. The generated scenes can be easily exported as Gaussian Splats or meshes directly into NVIDIA Isaac Sim for robotic simulation
An open-source engine based on Three.js for displaying Gaussian Splatting 3D worlds with over 100 million vertices in any browser. Accessible on desktop, mobile, and VR headsets (no plugins or installation required)
Create infinite interactive worlds that can be explored in real time with the Genie 3 model. It generates dynamic simulations with realistic physics from text or images. Currently available only to Google AI Ultra US subscribers
An open-source world simulator running at 16 FPS with real-time keyboard/mouse interaction and latency <1 s. Generation from a single photo with lasting consistency for embodied AI
An open-source model that simulates interactive worlds in real time and can be controlled via text, mouse, and keyboard. This AI simulator can run on a consumer gaming PC
Generate realistic videos or simulations of environments and agents: Dreamer4 uses deep reinforcement learning to explore and learn from new virtual worlds
A simulation created by researchers in which autonomous AIs live together. Activities, social interaction, etc.