AnyGroundBench - Industry TVG: leaderboard
Metric: tIoU@0.3 (%; share of queries whose temporal IoU is at least 0.3) on the 169 industry-domain test queries (MECCANO and ENIGMA industrial assembly videos), temporal video grounding (the queried event's start and end), zero-shot, default inference settings (Qwen3.5 with thinking disabled); higher is better. Source: arxiv.org. Saturation forecast: Around April 2028. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 39.6 |
| 2 | Gemini 3 Flash | 30.7 |
| 3 | Gemini 3.1 Pro (Preview) | 21.8 |
| 4 | Gemini 2.5 Flash | 21.3 |
| 5 | Qwen 3.5 9B (Non-reasoning) | 17.1 |
| 6 | Qwen 3.5 4B (Non-reasoning) | 14.7 |
| 7 | GPT-5.1 | 9.46 |
| 8 | GPT-4o | 8.87 |
| 9 | InternVL3-8B | 5.91 |
| 10 | InternVL3-14B | 5.32 |
| 11 | InternVL3.5-8B | 2.95 |
Interactive version: theaggregate.ai/benchmark?slug=anygroundbench-industry-tvg · How It Works · Data refreshed daily, snapshot 2026-09-29.