DeepYardDeepYard
S

SimWorlds

Generate dynamic 4D physics worlds from text for embodied AI training and simulation

Open SourceFree

About

Multi-agent system that transforms natural language descriptions into fully simulated 4D scenes with realistic physics. Handles liquids, particles, rigid bodies, and articulated mechanisms to create training environments for embodied AI agents. Designed for researchers building physically grounded agents that need to interact with complex, dynamic worlds.

Details

Typecoding-agent
Deploymentself-hosted
Supported Models

Tags

multi-agentautonomousopen-sourcepythonframework