WildTableBench - Aggregate Verification: leaderboard
A WildTableBench (2026) subtype on judging a claimed sum, average or count true or false, asked about real-world table images from the web. Accuracy in percent, free-form answers judged against the reference by GPT-5.2; higher is better.
Metric: Accuracy (%) on the Aggregate Verification subtype (C3-A), 928 human-verified questions over 402 real-world table images, free-form answers judged by GPT-5.2 against the reference answer, high reasoning effort where configurable; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 21 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Kimi K2.5 | Thinking | 81.8 | #272 |
| 2 | Seed 2.0 Pro | High | 63.6 | #226 |
| 3 | Gemini 3 Flash | 54.5 | #180 | |
| 4 | Gemini 3 Pro | 54.5 | #146 | |
| 5 | GPT-5.2 | High | 45.5 | #190 |
| 6 | Claude Opus 4.6 | High | 45.5 | #114 |
| 7 | Qwen 3 VL 235B A22B Instruct | 36.4 | #541 | |
| 8 | Qwen 3 VL 32B Instruct | 36.4 | #607 | |
| 9 | Qwen 3 VL 235B A22B (Thinking) | 36.4 | #462 | |
| 10 | Claude Sonnet 4.6 | High | 36.4 | #179 |
| 11 | GLM-4.6V | 27.3 | #721 | |
| 12 | GPT-5 Mini | High | 27.3 | #373 |
| 13 | Qwen 3 VL 8B Instruct | 18.2 | #1043 | |
| 14 | Qwen 3 VL 8B (Thinking) | 18.2 | #924 | |
| 15 | Qwen 3 VL 32B (Thinking) | 18.2 | #577 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=wildtablebench-aggregate-verification · How It Works · Data refreshed daily, snapshot 2026-11-22.