ClinSQL: leaderboard
Metric: Execution Score (%). Source: arxiv.org. Saturation forecast: Around December 2026. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 Mini | 74.67 |
| 2 | Gemini 2.5 Pro | 73.73 |
| 3 | DeepSeek R1 0528 | 69.15 |
| 4 | GPT-5 Chat | 68.42 |
| 5 | GPT-4.1 | 67.79 |
| 6 | Gemini 2.5 Flash | 65.01 |
| 7 | DeepSeek V3.1 | 60.71 |
| 8 | O4 Mini (2025-04-16) | 59.22 |
| 9 | Qwen 3 235B A22B 2507 Instruct | 58.63 |
| 10 | Grok 4 Fast (Reasoning) | 58.46 |
| 11 | Qwen 3 Coder 480B A35B Instruct | 58.18 |
| 12 | GPT-5 Nano | 52.03 |
| 13 | Qwen 3 Next 80B A3B Instruct | 49.26 |
| 14 | Qwen 3 235B A22B 2507 FP8 (Thinking) | 48.54 |
| 15 | Llama 4 Maverick Instruct | 47.49 |
Interactive version: theaggregate.ai/benchmark?slug=clinsql · How It Works · Data refreshed daily, snapshot 2026-09-25.