Humanities data-extraction benchmarks (RISE, University of Basel) run against UCSB inference services. Every cell is normalized to a 0-100 scale, so columns are directly comparable.
| # | Model | Runs | Avg 0-100 | Book Advert XML Text XML ยท Fuzzy | Bibliographic Data Printed List ยท Fuzzy | Fraktur Adverts Fraktur OCR ยท CER | Last Run |
|---|---|---|---|---|---|---|---|
| ๐ฅ | 3 | 93.2 | Sep 14, 2026 | ||||
| ๐ฅ | 3 | 88.3 | Sep 14, 2026 | ||||
| ๐ฅ | 3 | 87.4 | Sep 14, 2026 | ||||
| 4 | 3 | 87.0 | Sep 14, 2026 | ||||
| 5 | 3 | 78.7 | Sep 14, 2026 | ||||
| 6 | 3 | 74.1 | Sep 14, 2026 | ||||
| 7 | 1 | 96.1 | โ | โ | Sep 14, 2026 | ||
| 8 | 1 | 94.6 | โ | โ | Sep 15, 2026 | ||
| 9 | 1 | 94.2 | โ | โ | Sep 14, 2026 | ||
| 10 | 1 | 94.1 | โ | โ | Sep 14, 2026 | ||
| 11 | 1 | 93.1 | โ | โ | Sep 14, 2026 | ||
| 12 | 1 | 91.2 | โ | โ | Sep 15, 2026 | ||
| 13 | 1 | 89.9 | โ | โ | Sep 15, 2026 | ||
| 14 | 1 | 87.7 | โ | โ | Sep 15, 2026 | ||
| 15 | 1 | 79.1 | โ | โ | Sep 15, 2026 | ||
| 16 | 1 | 78.5 | โ | โ | Sep 15, 2026 |