Own the stack
C.A.T. is built from scratch in C++20 with Vulkan, an ECS architecture, explicit system lifecycles, and a custom asset pipeline. We learn by making the foundations visible.
An independent game company and research studio building a custom engine, playable worlds, and tools for understanding behavior.
Cat Game Research exists to make interactive worlds where technology serves imagination. We build the engine, the tools, and the game together so that a design idea can travel all the way from a question to a playable experience.
Our work sits at the intersection of game development, low-level graphics, behavioral intelligence, and formal research. The goal is not to make technology impressive for its own sake. The goal is to make systems that give players richer choices, more legible consequences, and worlds that feel alive.
C.A.T. is built from scratch in C++20 with Vulkan, an ECS architecture, explicit system lifecycles, and a custom asset pipeline. We learn by making the foundations visible.
Physics, rendering, VFX, dialogue, AI, and persistence are not isolated demos. They are ingredients for games where player decisions can matter beyond a scripted beat.
Our papers, textbook, APIs, benchmarks, lab tools, and Frontier transmissions document the decisions behind the work so other builders can inspect, learn, and challenge them.
Our vision is to grow an independent studio capable of delivering AAA-quality craft without losing the curiosity of a research lab. That means ambitious worlds, strong art direction, robust engineering, and respect for the player’s time and intelligence.
“AAA” describes the quality bar we are pursuing, not a claim about our current scale, funding, headcount, or shipped catalogue. Today, Cat Game Research is a focused, evolving program with working prototypes, research artifacts, and a public development network.
Current state: independent / building / publishingWe design for discovery, agency, and emotional memory. Complexity should produce meaningful possibility, not friction for its own sake.
We distinguish implemented systems from experiments and roadmaps. Performance, determinism, security, and limitations belong in the story.
We protect unusual ideas and let engineering expand the design space. A small studio can take risks when it keeps its tools understandable and its scope disciplined.
We share working knowledge through documentation, formal papers, benchmarks, and public tools. Research is more useful when others can reproduce the path.
The Cat Game Research network is the public surface of the studio: play the experiments, inspect the APIs, read the papers, explore the behavioral lab, follow the development log, and see how ideas become systems.