DDx-TRACE - Workup Order: leaderboard

Metric: Workup order concordance (0-1): share of comparable pairs of requested exams asked in the physician-labelled stage order on DDx-TRACE (211 physician-adjudicated multimodal neuroradiology cases), the model starts from limited history, requests imaging in free form (at most 6 requests), updates a four-item probabilistic differential after each turn and stops with a localized diagnosis, provider default decoding; endpoints judged against case rubrics by Gemini 3 Flash at temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked.

Top models

#ModelScore
1Gemini 3 Flash0.62
2Gemini 3.1 Pro (Preview)0.57
3GPT-5.40.57
4Claude Sonnet 4.60.55
5Lingshu-32B0.51
6MedGemma-27B-IT0.44
7Gemma 3 27B0.41
8Qwen 3.5 4B0.35
9Qwen 3.5 9B0.32
10Qwen 3.5 27B0.32
11Qwen 3.5 35B A3B0.19
12GPT-5.4 Mini0.18
13MedGemma 1.5 4B0.1

Interactive version: theaggregate.ai/benchmark?slug=ddx-trace-workup-order · How It Works · Data refreshed daily, snapshot 2026-10-07.