X-PCR: 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 averaged over the six stages, from image quality assessment to clinical decision-making, in percent; higher is better.
Metric: Stage-wise accuracy (%) averaged over the six stages of the progressive clinical reasoning chain (image quality, anatomy, lesions, diagnosis, severity grading, clinical decision), 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 February 2027. 21 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 76.24 | #178 |
| 2 | InternVL3-32B | 70.14 | |
| 3 | Gemini 2.5 Pro | 69.61 | #276 |
| 4 | Qwen 2.5 VL 72B | 69.15 | #781 |
| 5 | GLM-4.5V | 67.57 | #754 |
| 6 | Qwen 3 VL 30B A3B | 67.09 | #784 |
| 7 | GPT-5 Nano | 66.9 | #906 |
| 8 | Gemini 2.5 Flash | 65.8 | #463 |
| 9 | Qwen 2.5 VL 7B | 65.54 | #1862 |
| 10 | MedGemma-27B-IT | 65.21 | #1350 |
| 11 | Qwen 3 VL 8B | 64.72 | #917 |
| 12 | InternVL3-8B | 64.6 | #1782 |
| 13 | Claude Haiku 4.5 | 61.84 | #583 |
| 14 | Lingshu-7B | 61.66 | #1635 |
| 15 | HuaTuoGPT-Vision-34B | 57.37 | #1828 |
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 · How It Works · Data refreshed daily, snapshot 2026-11-21.