WildTableBench - Verification: leaderboard
Accuracy in percent on the verification category: true or false claims about cell values, aggregates and filtered counts, free-form questions about real-world table images collected from the web (WildTableBench, 2026), answers judged against the reference by GPT-5.2. Higher is better.
Metric: Accuracy (%) on the Verification question category (C3, question-weighted over its subtypes), 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 | Gemini 3 Pro | 71.6 | #145 | |
| 2 | Seed 2.0 Pro | High | 68.7 | #222 |
| 3 | Kimi K2.5 | Thinking | 67.2 | #266 |
| 4 | Claude Opus 4.6 | High | 61.2 | #113 |
| 5 | Gemini 3 Flash | 58.2 | #175 | |
| 6 | GPT-5.2 | High | 56.7 | #185 |
| 7 | Claude Sonnet 4.6 | High | 49.3 | #173 |
| 8 | Qwen 3 VL 32B Instruct | 41.8 | #606 | |
| 9 | Qwen 3 VL 235B A22B (Thinking) | 41.8 | #459 | |
| 10 | GPT-5 Mini | High | 41.8 | #364 |
| 11 | Qwen 3 VL 32B (Thinking) | 40.3 | #573 | |
| 12 | Qwen 3 VL 235B A22B Instruct | 34.3 | #532 | |
| 13 | GLM-4.6V | 34.3 | #710 | |
| 14 | O3 | High | 32.8 | #237 |
| 15 | Qwen 3 VL 8B (Thinking) | 28.4 | #910 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=wildtablebench-verification · How It Works · Data refreshed daily, snapshot 2026-11-21.