LiveClin: leaderboard
Metric: Case accuracy (%): share of the 1,407 first-half-2025 clinical cases (6,605 sequential ten-option questions generated from PubMed Central case reports and verified by physicians) in which every question is answered correctly, the conversation history kept across a case's questions; zero-shot, temperature 0 or the official setting for reasoning modes; higher is better. Source: arxiv.org. Saturation forecast: Around May 2027. 26 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O3 | 35.7 | #121 |
| 2 | GPT-5 | 35.5 | #91 |
| 3 | Gemini 2.5 Pro | 33.7 | #145 |
| 4 | GPT-4.1 | 31.8 | #240 |
| 5 | Claude 3.5 Sonnet (20241022) | 27.3 | #286 |
| 6 | GPT-4.1 Mini | 26.1 | #346 |
| 7 | Claude 3.7 Sonnet (20250219) (Thinking) | 25.6 | #196 (Claude 3.7 Sonnet (20250219)) |
| 8 | Gemini 2.0 Flash | 21.4 | #331 |
| 9 | Gemini 2.5 Flash | 19 | #237 |
| 10 | Qwen 2.5 VL 72B Instruct | 18.6 | #364 |
| 11 | Gemini 1.5 Pro | 17.7 | #442 |
| 12 | Qwen 2.5 VL 32B Instruct | 17.2 | #443 |
| 13 | Claude 3.5 Haiku (20241022) | 15.3 | #574 |
| 14 | GPT-4o Mini (2024-07-18) | 14.8 | #546 |
| 15 | GPT-4o (2024-11-20) | 14.7 | #369 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=liveclin · How It Works · Data refreshed daily, snapshot 2026-10-11.