LLM Stats (TVBench): leaderboard
LLM Stats mirror of TVBench, testing temporal reasoning about actions and events in video.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around February 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Seed 2.1 Pro | 80.5 | #73 |
| 2 | Qwen 3.7 Plus | 78.2 | #98 |
| 3 | Seed 2.1 Turbo | 77.2 | #93 |
| 4 | Seed 1.8 | 71.5 | #143 |
| 5 | Seed 2.0 Mini | 70.5 | #279 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-tvbench · How It Works · Data refreshed daily, snapshot 2026-11-05.