ClinSQL - Patient Demographics and Admissions: leaderboard
ClinSQL's Execution Score (a GPT-5 judge checks the executed result table against an expert rubric tree; percent, higher is better) restricted to the test tasks of its patient demographics and admissions scenario: cohort questions over demographic and administrative data such as admission type, readmission and length of stay. Queries come from chain-of-thought prompting with up to two execution-error refinements; Yale University and University of Washington, 2026.
Metric: Execution Score (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 22 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 Mini (2025-08-07) | 80.7 | #341 |
| 2 | Gemini 2.5 Pro | 79.22 | #262 |
| 3 | GPT-4.1 | 77.94 | #492 |
| 4 | DeepSeek R1 0528 | 74.86 | #425 |
| 5 | Qwen 3 Coder 480B A35B Instruct | 71.97 | #641 |
| 6 | Gemini 2.5 Flash | 70.53 | #452 |
| 7 | Qwen 3 235B A22B 2507 Instruct | 69.72 | #619 |
| 8 | O4 Mini (2025-04-16) | 67.35 | #362 |
| 9 | Grok 4 Fast (Reasoning) | 65.9 | #438 (Grok 4 Fast) |
| 10 | DeepSeek V3.1 | 65.66 | #500 |
| 11 | Qwen 3 Next 80B A3B Instruct | 59.68 | #763 |
| 12 | Qwen 3 235B A22B 2507 FP8 (Thinking) | 59.19 | #431 |
| 13 | Llama 4 Maverick Instruct | 55.67 | #1125 |
| 14 | Mistral Medium 3 | 51.97 | #1079 |
| 15 | GPT-5 Nano | 49.51 | #888 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=clinsql-patient-demographics-and-admissions · How It Works · Data refreshed daily, snapshot 2026-11-12.