PulseBench-Tab: leaderboard
Multilingual table-extraction benchmark over 1,820 human-annotated tables from document images in nine languages, scored by the T-LAG structural and content fidelity metric.
Metric: T-LAG Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5.1 | 89.04 | #9 |
| 2 | Gemini 3.8 Flash | 88.57 | #21 |
| 3 | GPT-6 Astra | 88.52 | #6 |
| 4 | GPT-5.6 Sol | 88.08 | #19 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=pulsebench-tab · How It Works · Data refreshed daily, snapshot 2026-10-20.