SceneActBench - Reconstruction: leaderboard
Metric: Task score (0-100): 100 3D-FRONT rooms from about 11 calibrated views: build the furnished scene from an empty Blender scene, scored by F-score at 5%; every configuration drives one headless Blender through the same MCP tool interface (scene and object inspection, Python execution, rendering) with task step budgets; each native geometric error or F-score is mapped to a 0 to 100 case score against a fixed reference, invalid outputs score 0, cases are averaged per task. Source: arxiv.org. Saturation forecast: Around 2032. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 (High) | 12.3 |
| 2 | Claude Sonnet 5 (High) | 10.5 |
| 3 | GPT-5.4 (Medium) | 10.4 |
| 4 | Claude Opus 4.6 (High) | 9.8 |
| 5 | MiniMax M3 (High) | 9.6 |
| 6 | Seed 2.0 Pro (High) | 8.8 |
| 7 | Step 3.7 Flash (High) | 8.6 |
| 8 | Kimi K2.6 (Thinking) | 8.5 |
| 9 | Gemini 3.1 Pro (Preview) (High) | 7.1 |
Interactive version: theaggregate.ai/benchmark?slug=sceneactbench-reconstruction · How It Works · Data refreshed daily, snapshot 2026-09-29.