WildTableBench - Value Modification: leaderboard
A WildTableBench (2026) subtype on recomputing an aggregate after specified cell values change, 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 Value Modification subtype (C4-M), 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 | GPT-5.2 | High | 66.7 | #185 |
| 2 | Gemini 3 Pro | 61.9 | #145 | |
| 3 | Gemini 3 Flash | 57.1 | #175 | |
| 4 | Kimi K2.5 | Thinking | 52.4 | #266 |
| 5 | Qwen 3 VL 235B A22B (Thinking) | 42.9 | #459 | |
| 6 | Claude Sonnet 4.6 | High | 42.9 | #173 |
| 7 | Seed 2.0 Pro | High | 42.9 | #222 |
| 8 | Qwen 3 VL 32B Instruct | 33.3 | #606 | |
| 9 | Claude Opus 4.6 | High | 33.3 | #113 |
| 10 | GPT-5 Mini | High | 28.6 | #364 |
| 11 | Qwen 3 VL 235B A22B Instruct | 23.8 | #532 | |
| 12 | GLM-4.6V | 23.8 | #710 | |
| 13 | O3 | High | 9.5 | #237 |
| 14 | Qwen 3 VL 4B Instruct | 4.8 | #1332 | |
| 15 | Qwen 3 VL 32B (Thinking) | 4.8 | #573 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=wildtablebench-value-modification · How It Works · Data refreshed daily, snapshot 2026-11-21.