TableVision - Ranking: leaderboard

TableVision (2026) category of questions over images of hierarchical tables: L2 reasoning, ranking located values (47 questions). Exact-match accuracy (%), zero-shot; higher is better.

Metric: Exact-match accuracy (%) on L2 reasoning, ranking located values (47 questions), zero-shot, the model answers directly from the table image; TableVision test set; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Glyph72.3
2Qwen 3 VL 8B Instruct70.5#1029
3Ovis2.5-9B70.2#1225
4GLM-4.1V-9B70.2#1125
5MiniCPM-V-2.659.6#3267
6HiPPO53.2
7DeepSeek-VL246.8#2927
8LLaVA-OneVision-7B36.2#2981
9InternVL3-8B31.9#1782
10Phi-3.5-Vision Instruct19.1#6268

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

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