LingBot 2.0: Open-Source World Model
kimmonismus · x · 2026-07-11
Robbyant has released the new open-source interactive world model LingBot-World 2.0, focusing on long-term stable real-time generation and exploration. The author claims it runs in real-time at 720p, 60fps, showing no significant degradation during a test of 20 scenes over a continuous 60-minute run, specifically solving the previous issue where "the world looks great for the first few seconds and then gradually breaks down."
This version also incorporates an agent loop: a VLM proposes events while the player is exploring, and a director agent injects new tasks and changes, allowing the world to continuously evolve rather than remaining static. The text mentions support for movement, combat, on-demand weather, and even multiplayer in a single shared world.
The project has open-sourced both model weights and code:
- 14B version is now available
- The paper also describes a 1.3B version designed for deployment on a single consumer-grade GPU
- An online playable demo hosted by Reactor is also available
Related event: Ant Group Open-Sources LingBot-World 2.0 Interactive World Model(15 posts)→
More from Multimodal
- HeyGen adds a media-sourcing skill for coding agents with 75k images and 10k tracks — HeyGen · 2026-07-22
- Reddit user seeks ComfyUI NSFW text-to-image and image-to-video workflows under 20 GB VRAM — hobbyist2020 · 2026-07-22
- Krea 2 users recommend a two-pass Clownshark sampler setup for sharper image details — listopalafoto · 2026-07-22
- Gemini Omni Flash turns a boat cabin into a cave in Flow by Google — chrisfirst · 2026-07-22
- A simple workflow to turn a photo into an image prompt using Gemini, Grok, or GPT Image — harshitagu72595 · 2026-07-22
- A Reddit user proposes a consistency LoRA to keep anime and game scenes visually stable — ThirdWorldBoy21 · 2026-07-22