Taplio - Build a growth routine on LinkedIn
Mini toolsDocVQA
Document VQA over scanned business documents. Tests OCR-grounded reading.
Multimodal
Multimodal
accuracy
Max 100.0%
Released Jul 2020
5
Results
5
Models scored
94.4%
Top: Llama 4 Maverick
93.3%
Median
Best results
Top primary scores; one row per model.
Frontier over time
Each dot is one model result; the line traces the running best score.
All results
Showing all configurations including non-primary alternates. · Show only primary
| # | Model | Score | Conditions | Eval date | Source | Flags |
|---|---|---|---|---|---|---|
| 1 | Claude Sonnet 3.5 | 95.2% | 0-shot · standard | Jun 20, 2024 | self reported | |
| 2 | Llama 4 Maverick | 94.4% | — | Apr 5, 2025 | self reported | primary |
| 3 | Llama 4 Scout | 94.4% | — | Apr 5, 2025 | self reported | primary |
| 4 | Pixtral Large | 93.3% | — | Nov 18, 2024 | self reported | primary |
| 5 | GPT-4o | 92.8% | — | Apr 16, 2025 | self reported | primary |
| 6 | Claude Haiku 3 | 88.8% | 0-shot · standard | Mar 4, 2024 | self reported | |
| 7 | Pixtral 12B | 78.6% | — | Oct 10, 2024 | self reported | primary |
