GameCraft-Bench - Art and Presentation: leaderboard
Metric: Art and Presentation score (%): rubric score for art and presentation (coherent visual style and polish), on the 140 GameCraft-Bench tasks (15 game families; complete Godot projects built from natural-language game specifications): a project that fails to launch or submits no parsable demonstration trace scores zero; otherwise the verifier replays the agent's mouse and keyboard traces into gameplay video and frames and a GPT-5.5 multimodal judge scores them against hidden rubric items; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 (High) | 36.86 |
| 2 | GPT-5.5 (High) | 32.94 |
| 3 | Kimi K2.6 | 27.99 |
| 4 | MiMo-V2.5-Pro | 20.65 |
Interactive version: theaggregate.ai/benchmark?slug=gamecraft-bench-art-and-presentation · How It Works · Data refreshed daily, snapshot 2026-09-29.