LingBot-World 2.0
An interactive video diffusion world model with 1.3 billion parameters that autoregressively predicts subsequent frames from an initial frame and action instructions, generating controllable virtual environments. It is classified as a lightweight world model.
Timeline (last 90 days)
A developer increased the inference speed of the LingBot-World 2.0 1.3B model on a single RTX 5090 from 6 FPS to 16 FPS and open-sourced the code.
The optimized LingBot-World 2.0 inference speed is 2.5 times faster than SGLang Diffusion.
The optimized LingBot-World 2.0 inference speed is 1.9 times faster than NVIDIA FlashDreams.
When mainstream inference engines run this model on the same hardware, the speed is generally stuck in the 6–8 FPS range.
The project's resolution is limited to 832×464.
The project currently only supports Linux, and the experience can be started by entering the command lingbot play dragon in the command line.
Users can upload an image along with a text prompt to begin predicting subsequent frames of the world, and use the keyboard AWSD keys and mouse to move and navigate within the generated world.
The author estimates that an RTX 4090 can reach approximately 12 FPS, but this has not yet been verified through actual testing.
All Facts (8)
Users can upload an image along with a text prompt to begin predicting subsequent frames of the world, and use the keyboard AWSD keys and mouse to move and navigate within the generated world.
50%UnverifiedThe optimized LingBot-World 2.0 inference speed is 2.5 times faster than SGLang Diffusion.
50%UnverifiedThe optimized LingBot-World 2.0 inference speed is 1.9 times faster than NVIDIA FlashDreams.
50%UnverifiedWhen mainstream inference engines run this model on the same hardware, the speed is generally stuck in the 6–8 FPS range.
50%UnverifiedThe project's resolution is limited to 832×464.
50%UnverifiedThe project currently only supports Linux, and the experience can be started by entering the command lingbot play dragon in the command line.
50%UnverifiedThe author estimates that an RTX 4090 can reach approximately 12 FPS, but this has not yet been verified through actual testing.
50%UnverifiedA developer increased the inference speed of the LingBot-World 2.0 1.3B model on a single RTX 5090 from 6 FPS to 16 FPS and open-sourced the code.
50%