ParseBench - Tables: leaderboard
LlamaIndex's ParseBench scores document parsing of human-verified enterprise pages from insurance, finance and government documents with rule-based checks; VLM rows only, parser products excluded. This board reports the Tables dimension: structural fidelity of tables with merged cells and hierarchical headers (GTRM, combining GriTS and TableRecordMatch).
Metric: GTRM Score (GriTS + TableRecordMatch). Source: parsebench.ai. Saturation forecast: Around February 2027. 121 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 (High) | 94.26 | #3 (Claude Opus 5.5) |
| 2 | Claude Opus 5.5 | 93.88 | #3 |
| 3 | GPT-6 Astra (Low) | 93.18 | #5 (GPT-6 Astra) |
| 4 | GPT-6.1 Terra (Medium) | 93.07 | #14 (GPT-6.1 Terra) |
| 5 | Claude Opus 5.5 (Low) | 92.71 | #3 (Claude Opus 5.5) |
| 6 | GPT-6 Sol (High) | 92.65 | #17 (GPT-6 Sol) |
| 7 | GPT-6.1 Sol (Low) | 92.46 | #6 (GPT-6.1 Sol) |
| 8 | GPT-6.1 Terra (Low) | 92.18 | #14 (GPT-6.1 Terra) |
| 9 | Claude Fable 5.1 | 91.53 | #10 |
| 10 | GPT-5.6 Sol (High) | 91.41 | #22 (GPT-5.6 Sol) |
| 11 | Gemini 3 Flash (High) | 91.19 | #105 (Gemini 3 Flash) |
| 12 | Gemini 3.5 Flash (Medium) | 91.15 | #68 (Gemini 3.5 Flash) |
| 13 | Gemini 3.1 Pro (Preview) | 91.13 | #66 |
| 14 | Claude Sonnet 5.5 | 91.07 | #13 |
| 15 | Gemini 3 Flash (Medium) | 91.03 | #105 (Gemini 3 Flash) |
No result here: #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=parsebench-tables · How It Works · Data refreshed daily, snapshot 2026-11-07.