Infinite Worlds with Versatile Interactions

2026-07-08

AI Summary

This work proposes a text-driven virtual world simulation system designed to deliver immersive experiences with unlimited duration, high responsiveness, and rich interactivity. Leveraging a causally pretrained language model and knowledge distillation, the system employs a collaborative architecture comprising a 14B primary model and a 1.3B lightweight deployment model. It introduces a novel dual-agent framework—featuring a pilot agent and a director agent—that unifies planning-based character behaviors with dynamic environment generation within a coherent world model. The system achieves stable 720p/60fps video streaming on a single GPU, supports synchronous multi-user interaction, and enables diverse actions such as attacking, archery, and spellcasting. To the best of our knowledge, this is the first system to realize high-quality, unbounded, and shared multi-user text-to-virtual-world generation.

Abstract

We present LingBot-World 2.0 (also known as LingBot-World-Infinity), an advanced iteration of LingBot-World featuring four distinct upgrades. (1) Our model achieves an unbounded interaction horizon while maintaining consistent output quality, benefiting from a carefully crafted causal pretraining paradigm. (2) Through distilling a real-time variant from the base model, our system guarantees rapid response time, sufficient to drive 720p video streams at 60 fps. (3) Compared to the previous version, this update introduces highly diverse interactive elements, comprising a broader spectrum of actions (e.g., attacking, archery, spell-casting, and shooting) alongside a richer variety of text-driven events. (4) We pioneer the integration of an agentic harness within the domain of world modeling, wherein a pilot agent is tasked with planning and executing character behaviors, while a director agent is responsible for synthesizing novel environmental elements as the scene progresses. Additionally, to facilitate a shared experience, we develop an interface that permits multiple players to simultaneously immerse themselves in this vivid world simulator. We pair our primary 14B model with a lightweight 1.3B counterpart, which supports effortless deployment on a single GPU.

Problem

• world modeling
• interactive simulation
• text-driven environment
• multi-agent system
• real-time interaction

Innovation

• causal pretraining
• real-time distillation
• text-driven interaction
• agentic harness
• world modeling