Maternity Case Record Digitisation - Discrete Fields Weighted F1: leaderboard
Metric: Support-weighted F1 (from 0-1, times 100) over the discrete selection fields treated as multi-class classification with null as its own class, on the 49 real redacted Maternity Case Records of the UBOMI BUHLE Pregnancy Exposure Registry (South Africa), one form image per call with the optimised extraction prompt and JSON template, temperature 0, two attempts at valid JSON; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 91 |
| 2 | GPT-5.4 | 89 |
| 3 | Gemini 2.5 Flash | 87 |
| 4 | Gemini 3.1 Flash Lite | 87 |
| 5 | Gemini 2.5 Pro | 85 |
| 6 | Claude Sonnet 4.6 | 84 |
| 7 | Claude Opus 4.6 | 83 |
| 8 | Claude Haiku 4.5 | 75 |
| 9 | GPT-5 Mini | 40 |
Interactive version: theaggregate.ai/benchmark?slug=maternity-case-record-digitisation-discrete-fields-weighted-f1 · How It Works · Data refreshed daily, snapshot 2026-10-07.