AI 做游戏终于不丑了：灰块变大片，角色动作全保留
AI 现在能几分钟搭出能玩的游戏，但画面像程序员美术：灰块、几何体、色块。这篇让 AI 把这种粗糙画面直接重绘成高质量成品——角色、动作、特效全都有，而且布局、镜头、每个角色的位置和出场时机一点不乱。它吃的是普通视频，不需要引擎底层数据，灰块、色块、线框都能当输入。做法是微调一个 33B 的视频生成模型，用 38 个 AI 做的游戏加虚幻引擎的模拟数据训练，还专门测了「实体对应」:该出现时出现、该消失时消失、数量不错。对比原模型，时间匹配率从 50.9% 提到 62.9%，漏掉的次数变化从 34.7% 降到 21.3%。这不是你明天能用的工具，但它是「AI 做游戏」从能玩到能看的转折点——下一步，AI 生成游戏可能就不只是原型，而是直接能上架的画面。
📄 原文摘要(英文)
AI coding agents can now build small playable games in minutes, but the results usually look like programmer art: gray blockouts, colored geometric primitives, and flat semantic blobs. We present AnyGameRenderer, a video-to-video generative renderer that turns such footage into high-quality visuals with characters, motion detail, and effects. Throughout, it keeps the layout, camera, and every entity's position and timing intact. Our AnyGameRenderer generalizes across a wide range of genres that agents build, such as tower defense, auto-battlers, top-down survivors, turn-based battles, kart racing, and crowd runners. AnyGameRenderer takes plain video as its game-state input, with no complex representations such as engine G-buffers, per-object class labels, or low-poly proxies. Gray blockouts, colored primitives, semantic maps, and wireframes all work as input. It also preserves entity correspondence: group counts stay correct, and entities appear and disappear on the right frame. We implement AnyGameRenderer by LoRA fine-tuning MiniMax-H3 ref2va, a 33B reference-video-conditioned video generation model. Training uses sim10k paired clips from three sources: (i) multi-stream rendering data from 38 agent-made games exported from GEAR, an AI game-creation tool; (ii) two-stream rendering data from Unreal Engine, with programmatically controlled crowds, count changes, and entity scales; and (iii) real gameplay recordings, with pairs reconstructed via 3D-box detection. To evaluate the entity correspondence, we propose an entity-level benchmark of 59 held-out cases with 753 entities, scored against ground-truth boxes. Compared with the base model, AnyGameRenderer better preserves entity correspondence: temporal match rises from 50.9\% to 62.9\%, missed count changes fall from 34.7\% to 21.3\%, and fewer entities are hallucinated or left unrendered.