TriViewBench - Level 3: leaderboard
Metric: Accuracy (%) over the 3,465 Level-3 questions (high object count, low occlusion) of all three reasoning categories; three orthographic views (front, side, top) of synthetic 3D tower scenes, direct prompting with a concise answer within 128 tokens, greedy decoding; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 81.9 |
| 2 | Claude 3.7 Sonnet | 67.99 |
| 3 | Qwen 3 VL 8B Instruct | 55.58 |
| 4 | Qwen 3 VL 32B Instruct | 54.26 |
| 5 | GPT-4o | 50.74 |
| 6 | Qwen 3 VL 4B Instruct | 43.61 |
| 7 | InternVL3.5-8B | 37.52 |
| 8 | Qwen 2.5 VL 32B Instruct | 35.3 |
| 9 | Qwen 2.5 VL 7B Instruct | 22.74 |
Interactive version: theaggregate.ai/benchmark?slug=triviewbench-level-3 · How It Works · Data refreshed daily, snapshot 2026-09-29.