TOC-Bench: leaderboard
TOC-Bench (2026) tests temporal object consistency in video LLMs: 2,323 human-verified questions over 1,951 videos from Charades, Perception Test, OVIS and MOSE, grounded in per-frame object tracks, on whether objects appear, recur, keep their identity, move and change state over time. Overall accuracy in percent, mostly with 32 sampled frames; higher is better.
Metric: Overall accuracy (%) on TOC-Bench (2,323 human-verified questions over 1,951 videos from Charades, Perception Test, OVIS and MOSE, grounded in per-frame object tracks and kept only when not answerable from text alone, single frames or shuffled frames), mostly 32 uniformly sampled frames, deterministic scoring of multiple-choice, statement-pair, numeric and ordering formats; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 21 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 47.1 | #56 |
| 2 | Kimi K2.6 | 45 | #197 |
| 3 | Gemini 3.1 Pro (Preview) | 43.9 | #102 |
| 4 | Grok 4.3 | 43.8 | #351 |
| 5 | Seed 2.0 Lite | 42 | #306 |
| 6 | GLM-5V Turbo | 37.9 | #365 |
| 7 | Qwen 2.5 VL 72B | 34 | #782 |
| 8 | Gemini 3.1 Flash Lite (Preview) | 33.8 | #419 |
| 9 | GPT-5.4 Mini | 33.1 | #409 |
| 10 | Qwen 3 VL 8B (Thinking) | 32.1 | #924 |
| 11 | MiMo-V2-Omni | 31.6 | #465 |
| 12 | InternVL3-8B | 30.1 | #1795 |
| 13 | VideoChat-R1.5-7B | 29.4 | #1260 |
| 14 | LLaVA-Video-72B | 28.9 | #1211 |
| 15 | VideoLLaMA3-7B | 28.4 | #2064 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=toc-bench · How It Works · Data refreshed daily, snapshot 2026-11-22.