WanToFight: Real-Time Generative Game Engine for Multi-Player Combat Interaction

2026-07-14

AI Summary

Existing generative game engines struggle to simultaneously support multi-player control, real-time inference, complex physical interactions, and adversarial gameplay. This work proposes the first generative game engine capable of enabling real-time two-player versus matches in *The King of Fighters '97*, built upon the Wan-1.3B video diffusion Transformer. The system delivers a complete competitive experience through streaming autoregressive generation, character binding, and keyboard injection mechanisms. Key technical innovations include block-wise causal attention, rolling KV caching, a vision-anchored player association module, gated local-causal keyboard injection, and model compression via DMD distillation and VAE decoder pruning. The resulting system achieves real-time gameplay generation at 512×384 resolution and 30 FPS on a single NVIDIA RTX 5090 GPU.

Abstract

We present WanToFight, a generative game engine that simulates real-time, two-player The King of Fighters '97 (KOF~'97) gameplay from keyboard input. Prior generative game engines target either single-player first-person settings or non-real-time cooperative scenarios; multi-player control, real-time inference, complex physical interaction, and adversarial gameplay have not been jointly addressed. WanToFight closes this gap with three components built on the Wan-1.3B video diffusion transformer: a streaming autoregressive generator with block-causal attention and a rolling KV cache; a visually grounded Player Association module that binds each player's keyboard signal to a character identity; and a gated, locally causal keyboard injection module trained with a single-player-to-full-gameplay curriculum. A four-step DMD-distilled student paired with a pruned VAE decoder sustains 30FPS at 512x384 on a single NVIDIA RTX 5090 over the duration of a complete match. To our knowledge, WanToFight is the first generative game engine to combine multi-player control, real-time inference, complex physical interaction, and adversarial gameplay in one system.

Problem

• multi-player control
• real-time inference
• complex physical interaction
• adversarial gameplay
• generative game engine

Innovation

• generative game engine
• real-time inference
• multi-player combat
• video diffusion transformer
• adversarial gameplay