Craftium: Bridging Flexibility and Efficiency for Rich 3D Single- and Multi-Agent Environments
Mikel Malag\'on, Josu Ceberio, Jose A. Lozano

TL;DR
Craftium is a flexible, customizable 3D environment platform that enables efficient creation of complex single- and multi-agent worlds, outperforming existing tools in computational efficiency.
Contribution
It introduces a novel platform that balances richness and efficiency for 3D environments, supporting multi-agent interactions and procedural generation.
Findings
Achieves +2000 steps/sec more than Minecraft-based frameworks.
Supports diverse environments from simple tasks to large worlds.
Reduces computational costs while maintaining environment complexity.
Abstract
Advances in large models, reinforcement learning, and open-endedness have accelerated progress toward autonomous agents that can learn and interact in the real world. To achieve this, flexible tools are needed to create rich, yet computationally efficient, environments. While scalable 2D environments fail to address key real-world challenges like 3D navigation and spatial reasoning, more complex 3D environments are computationally expensive and lack features like customizability and multi-agent support. This paper introduces Craftium, a highly customizable and easy-to-use platform for building rich 3D single- and multi-agent environments. We showcase environments of different complexity and nature: from single- and multi-agent tasks to vast worlds with many creatures and biomes, and customizable procedural task generators. Benchmarking shows that Craftium significantly reduces the…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
Taxonomy
TopicsOpen Source Software Innovations · Digital Mental Health Interventions · Reinforcement Learning in Robotics
