CraftBench-UE (C++, File-Shell): leaderboard
CraftBench-UE (2026) evaluates coding agents on bounded gameplay and editor tasks in Unreal Engine 5.8, reconstructing each saved submission in a fresh project and applying deterministic build, runtime and saved-asset checks with no LLM judge. This board is C++ source tasks completed within a 40-minute budget under a file-and-shell agent without an editor MCP server, over 33 C++ tasks, driven by the Claude Code CLI; the metric is task completion in percent, higher is better.
Metric: Task completion (%; C++ tasks completed within 40 minutes over 33 tasks, file and shell tools without editor MCP, deterministic build and runtime checks, Claude Code driver). Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 87.88 | #29 |
| 2 | Grok 4.6 | 84.85 | #44 |
| 3 | Claude Sonnet 5 | 81.82 | #108 |
| 4 | GLM-5.3 Flash | 78.79 | #119 |
| 5 | DeepSeek V4 Pro (0813) | 78.79 | #81 |
| 6 | Gemini 3.7 Flash | 69.7 | #32 |
| 7 | GPT-5.6 Luna | 63.64 | #117 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=craftbench-ue-c-plus-plus-file-shell · How It Works · Data refreshed daily, snapshot 2026-11-12.