TOC-Bench - Hallucination Diagnostic Accuracy: leaderboard
Hallucination diagnostic accuracy of TOC-Bench (2026), video questions grounded in per-frame object tracks and kept only when not answerable from text alone, single frames or shuffled frames: accuracy on question variants about a nonexistent subject or a nonexistent event, where the model must reject a plausible but visually unsupported option. Percent; higher is better.
Metric: Hallucination diagnostic accuracy (%, HDA): accuracy on the nonexistent-subject and nonexistent-event question variants, where the model must reject a plausible but visually unsupported object or event, 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 September 2028. 21 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4.3 | 53.8 | #351 |
| 2 | Kimi K2.6 | 52.9 | #197 |
| 3 | Gemini 3.1 Pro (Preview) | 47.7 | #102 |
| 4 | Qwen 3 VL 8B (Thinking) | 43.8 | #924 |
| 5 | GPT-5.5 | 42.8 | #56 |
| 6 | Qwen 2.5 VL 72B | 42.6 | #782 |
| 7 | Gemini 3.1 Flash Lite (Preview) | 41.2 | #419 |
| 8 | Seed 2.0 Lite | 41.2 | #306 |
| 9 | Qwen 2.5 VL 7B | 35.3 | #1878 |
| 10 | MiMo-V2-Omni | 35.1 | #465 |
| 11 | GLM-5V Turbo | 34.7 | #365 |
| 12 | VideoChat-R1.5-7B | 33.1 | #1260 |
| 13 | Video-R1 | 32.5 | #1384 |
| 14 | GPT-5.4 Mini | 28.4 | #409 |
| 15 | InternVL3-8B | 27.8 | #1795 |
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-hallucination-diagnostic-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-22.