Reportedly: Kunlun Tech releases Matrix-Game 4.0, the first video large model game running on a single consumer GPU, and open-sources it
Reported by IT Home and Pandaily · not yet confirmed by the company or a second independent outlet. We update this page when it is.
Kunlun Tech released Matrix-Game 4.0, the first video large model game running on a single consumer GPU, achieving 480p at 60 FPS with under 150 ms latency on an NVIDIA RTX 5090. The project is open-sourced on GitHub and Hugging Face, IT Home reported.
China context
- Original name
- 昆仑万维
- Outside China
- Open weights · ithome.com
- Claims
- Company-reported; not yet independently evaluated
- For builders
- Developers outside China can access the open-source code and model weights on GitHub and Hugging Face to test and build on Matrix-Game 4.0.
- For investors
- The open-source release may accelerate adoption and ecosystem development around Kunlun Tech's game AI technology, but revenue model and commercial traction are not yet evidenced.
Translated from Chinese. Quotes and facts link to the original sources.
Kunlun Tech announced Matrix-Game 4.0, a video large model game that runs on a single NVIDIA RTX 5090 consumer GPU. It delivers 480p resolution at 60 FPS with latency below 150 ms from key press to screen response. The system combines a game engine for world state and interaction rules with a video generation model for real-time rendering, enabling a closed loop from code-generated game logic to structural constraints to real-time video rendering. The project is open-sourced, with repositories on GitHub and Hugging Face. Future plans include a refiner for higher resolution, multi-player consistency across perspectives, finer-grained control, and support for more game types.
Matrix-Game 4.0 integrates a game engine with a video generation model to achieve real-time interactive rendering on consumer hardware. The reported specifications are 480p at 60 FPS with under 150 ms latency on an NVIDIA RTX 5090. The open-source release allows independent verification of these performance claims.
Game developers outside China can now access an open-source system that runs a video large model game on a single consumer GPU, potentially reducing the hardware barrier for prototyping AI-driven games. The availability of the code and model weights on GitHub and Hugging Face enables direct experimentation and benchmarking against existing approaches.
Kunlun Tech positions Matrix-Game 4.0 as a foundation for game prototyping, personalization, and user-generated content. The open-source strategy may attract developers and establish an ecosystem around its technology, but commercial terms and support are not detailed in the evidence.
The next verifiable signals are the release of the refiner model for higher resolution, demonstration of multi-player consistency, and independent benchmarks of latency and frame rate on the specified hardware. The open-source repository activity and community contributions will indicate adoption.