VGenST-Bench - Quantity Change Tracking: leaderboard
VGenST-Bench (2026) tests spatio-temporal reasoning of multimodal LLMs on short videos synthesized from validated scene graphs with text-to-image and image-to-video models. Quantity Change Tracking task: track the item counts on a workspace as items are added or removed (figural scale, egocentric, dynamic). Accuracy in percent on multiple-choice questions under circular evaluation (right under every rotation of the options), 8 sampled frames; higher is better.
Metric: Quantity Change Tracking (figural scale, egocentric, dynamic) accuracy (%) on the base multiple-choice questions of VGenST-Bench under circular evaluation (correct only if right under every cyclic permutation of the options), macro-averaged over the task's QA types, 8 uniformly sampled frames from about 100 synthesized videos per task (generated with text-to-image and image-to-video models from validated scene graphs); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 15 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Flash Lite | 71.6 | #371 |
| 2 | Gemini 3 Flash | 71.4 | #175 |
| 3 | GPT-5.4 | 71 | #135 |
| 4 | Gemma 4 31B (IT) | 66.9 | #350 |
| 5 | Gemma 4 26B A4B (IT) | 61.9 | #556 |
| 6 | Qwen 3.5 27B | 57 | #375 |
| 7 | InternVL3.5-30B-A3B | 55.7 | #1267 |
| 8 | Qwen 3.5 9B | 54.4 | #828 |
| 9 | Qwen 3.5 4B | 51.6 | #1241 |
| 10 | GPT-5.4 Nano | 48.2 | #651 |
| 11 | GPT-5.4 Mini | 46.3 | #406 |
| 12 | Kimi K2.6 | 45 | #192 |
| 13 | InternVL3.5-8B | 42 | #1604 |
| 14 | GLM-4.6V-Flash | 40 | #1040 |
| 15 | InternVL3.5-4B | 38.3 | #1930 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=vgenst-bench-quantity-change-tracking · How It Works · Data refreshed daily, snapshot 2026-11-21.