VisualTextTrap (TemporalBench): leaderboard

Text overlay-induced hallucination in video question answering on VisualTextTrap (2026), built on TemporalBench videos with on-screen text that agrees with, contradicts or is absent from the visual evidence. Overall multiple-choice accuracy in percent; higher is better.

Metric: Overall VQA accuracy (%) on the VisualTextTrap videos built on TemporalBench (text overlays rendered on every frame under text-free, text-congruent and text-contradictory conditions, the contradictory overlays written by Claude-Sonnet-4.6), multiple-choice video QA under the default configuration of each model; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 5 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3.1 Pro (Preview)62.5#99
2Qwen 3 VL 235B A22B41.6#485
3InternVL3.5-241B-A28B40.5#733
4Qwen 3 VL 30B A3B38.3#784
5Qwen 3 VL 8B Instruct24.2#1029

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=visualtexttrap-temporalbench · How It Works · Data refreshed daily, snapshot 2026-11-21.