MedRepBench - Value Recall (OCR-Assisted): leaderboard

Metric: Field-level recall (%; measured value of each ground-truth lab item, matched by exact normalized item name within the model's top-K items, K = number of reference items; VLM on the report image plus OCR text). Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.

Top models

#ModelScore
1Llama 4 Maverick87.35
2InternVL3-14B86.53
3InternVL3-38B85.64
4InternVL3-8B84.87
5Qwen 2.5 VL 7B82.46
6Llama 4 Scout78.86

Interactive version: theaggregate.ai/benchmark?slug=medrepbench-value-recall-ocr-assisted · How It Works · Data refreshed daily, snapshot 2026-09-26.