X-PCR - Clinical Decision-Making: leaderboard

X-PCR (2026) tests multimodal models on ophthalmic diagnosis as a six-stage reasoning chain over six imaging modalities, with multiple-choice questions built from 51 public datasets and checked by ophthalmologists. Accuracy at the clinical decision stage, whose questions ophthalmologists wrote, in percent; higher is better.

Metric: Accuracy (%) at stage 6, choosing management (items written by ophthalmologists), on X-PCR's ophthalmic multiple-choice VQA (26,415 images from 51 public datasets, six imaging modalities, 52 diseases), zero-shot, temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 21 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-554.71#178
2GPT-5 Nano48.92#906
3Qwen 2.5 VL 72B47.06#781
4GLM-4.5V46.58#754
5MedGemma-27B-IT45.32#1350
6Qwen 3 VL 30B A3B44.4#784
7Claude Haiku 4.543.83#583
8InternVL3-32B43.22
9Qwen 3 VL 8B41.49#917
10HuaTuoGPT-Vision-34B40.64#1828
11LLaVA-v1.5-13B39.37#6926
12Qwen 2.5 VL 7B38.31#1862
13LLaVA-Next-LLaMA3-8B38.27#5654
14Gemini 2.5 Pro37.35#276
15InternVL3-8B37.22#1782

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=x-pcr-clinical-decision-making · How It Works · Data refreshed daily, snapshot 2026-11-21.