跳到正文
arXiv cs.AI· Bizhe Bai, Jiliang Jian, Yuqing Yan, Tao Chen, Ying Wen·· 1 小时前AI 评分38

3D GameDev Arena:评测能规划、构建并试玩超长程浏览器游戏的 Coding Agent 基准

3D GameDev Arena: Benchmarking Coding Agents That Plan, Build, and Playtest extreme Long-Horizon Browser Games

AI 导读

研究者推出 3D GameDev Arena,这是首个针对从零构建 3D 浏览器游戏的极限长程基准,包含 14 个类型的 108 个任务。自动化的“规划—实现—测试—修复”流水线平均每次运行约 1,000 次 LLM 调用,最长一次使用 14,407 次调用、2.13B tokens 和 608 个素材。

正文 · 原文

View PDF HTML (experimental)

Abstract:Coding agents now resolve repository issues and generate compact playable games, but whether they can build a large 3D game from scratch remains open. Such games tightly couple mechanics, scenes, assets, and interaction logic, and their quality rests on player enjoyment, which existing benchmarks rarely assess. Thus, we introduce 3D GameDev Arena, a first extreme long-horizon benchmark of 108 from-scratch game-development tasks across 14 genres. Given only a text specification and UI/UX reference images, a fully automated plan--implement--test--repair pipeline builds each task into a self-contained browser game. Runs average $\sim$1,000 LLM calls ; the longest uses 14,407 calls, 2.13B tokens, and 608 assets. Evaluation spans (1) objective code-level functional acceptance against the game specification, (2) agent-based QA and playtesting of 3D presentation and playability, and (3) blind pairwise Bradley-Terry human preference in a public arena. The preliminary findings indicate that long-horizon 3D game development remains far from solved, and code-level correctness alone does not dictate player preference. The public arena is hosted online.
Comments: 23 Pages
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.24073 [cs.AI]
  (or arXiv:2611.24073v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.24073

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Bizhe Bai [view email]
[v1] Wed, 18 Nov 2026 09:03:53 UTC (16,933 KB)

来源:arXiv cs.AI · arxiv.org