LingBot Video 1.3B and LingBot-World-v2 14B land on ComfyUI with 8GB VRAM support
Two LingBot models—a 1.3B video generator and a 14B world-generation model—have been repackaged as ComfyUI custom nodes with GGUF quantization, enabling local video and scene synthesis on consumer GPUs.
LingBot Video 1.3B and LingBot-World-v2 14B now run on 8GB VRAM cards after being repackaged for ComfyUI with GGUF quantization support.
LingBot Video 1.3B is a text-to-video and image-to-video generator that has been integrated into ComfyUI as custom nodes. Community members report the architecture quantizes cleanly without major quality loss. The repackaging makes the model accessible to the ComfyUI ecosystem, which has become the standard for local generative workflows among practitioners running open-weight models. Text-to-video and image-to-video capabilities slot directly into existing ComfyUI pipelines, letting users chain video generation with upscaling, frame interpolation, and other post-processing nodes already in their libraries.
LingBot-World-v2 14B is a causal world-generation model that produces spatial environments and scenes. The repackaged version ships as a GGUF checkpoint optimized for memory efficiency, a format that trades some inference speed for dramatically lower VRAM consumption. World-generation models occupy a niche between pure text-to-image synthesis and full 3D scene construction—they output coherent spatial layouts and environmental context that can feed downstream rendering or game-engine workflows. The 14B parameter count puts it in the same weight class as mid-tier language models, but the causal architecture is tuned for spatial reasoning rather than conversational output.
Both integrations are available on GitHub under the RealRebelAI namespace. The LingBot Video nodes support text-to-video and image-to-video workflows, while the World-v2 nodes handle scene generation and spatial reasoning tasks. The 8GB VRAM floor puts both models within reach of RTX 3060 and 4060 Ti users running locally—a significant accessibility threshold for hobbyists and small studios that can't justify multi-GPU rigs. GGUF quantization has become the standard path for fitting larger models onto consumer hardware. The format originated in the llama.cpp ecosystem and has since spread to image and video models as practitioners push for local inference without cloud API costs. ComfyUI's node-based interface eliminates the need for standalone inference scripts and Python environment wrangling, lowering friction for casual users experimenting with open-weight releases.
