IDEAL-Bench - Object Recognition Recall: leaderboard
Object recognition on IDEAL-Bench (2026), a single-image 3D indoor layout benchmark: the percent of in-view ground-truth objects in non-grid rooms that a predicted object of the same category matches. Higher is better.
Metric: Recognition rate (%): share of in-view ground-truth objects in non-grid rooms that a same-category prediction matches, IDEAL-Bench 3D layout task; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 89.9 | #266 |
| 2 | Claude Sonnet 4.6 | 85.4 | #161 |
| 3 | GPT-5.4 | 82.2 | #123 |
| 4 | Gemma 4 31B (IT) | 81.8 | #342 |
| 5 | Qwen 3 VL 30B A3B Instruct | 79.9 | #871 |
| 6 | GLM-4.6V | 76.8 | #701 |
| 7 | Qwen 3 VL 235B A22B Instruct | 76.7 | #522 |
| 8 | Qwen 3 VL 8B Instruct | 73 | #981 |
| 9 | GPT-4o | 72.8 | #741 |
| 10 | Qwen 2.5 VL 72B Instruct | 70.6 | #846 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=ideal-bench-object-recognition-recall · How It Works · Data refreshed daily, snapshot 2026-11-13.