RxScribe Bench - Hard Fields (Delivered): leaderboard
RxScribe Bench (2026) robustness axis: field credit on the 1,740 fields that annotators marked as partly legible or blurred but still recoverable, with a skipped field counted as zero. Percent over three runs on each of 200 handwritten prescriptions; higher is better.
Metric: Hard-field score (%; field credit on the 1,740 fields annotated as partially legible or blurred but recoverable, with a skipped field scored as zero; 200 handwritten Indian outpatient prescription images, three independent cold runs per image, image and JSON schema only, deterministic field-by-field scoring against human ground truth with no model judge). Source: arxiv.org. Saturation forecast: Around 2030. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 51.03 | #91 |
| 2 | GPT-5.6 Sol | 49.23 | #29 |
| 3 | Muse Spark 1.2 | 48.08 | #97 |
| 4 | Claude Opus 5 | 46.35 | #20 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=rxscribe-bench-hard-fields-delivered · How It Works · Data refreshed daily, snapshot 2026-11-12.