LingxiDiagBench (Dynamic, Free-Form) - 2-Class Accuracy: leaderboard
Metric: Accuracy (%, 0-100) on binary depression-versus-anxiety classification (cases without comorbidity) after the consultation; LingxiDiagBench-Dynamic on LingxiDiag-16K cases: the model leads a multi-turn consultation with a Qwen3-32B patient agent built from the case EMR, then diagnoses; doctor strategy: free-form questioning as a senior psychiatrist; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4.1 Fast | 88.6 | #208 |
| 2 | Qwen 3 32B | 86.5 | #424 |
| 3 | GPT-5 Mini | 86.4 | #176 |
| 4 | DeepSeek V3.2 | 84.6 | #198 |
| 5 | Gemini 3 Flash | 84.1 | #93 |
| 6 | Claude Haiku 4.5 | 84.1 | #271 |
| 7 | Kimi K2 (Thinking) | 84.1 | #236 (Kimi K2) |
| 8 | GPT-OSS-20B | 80.8 | #499 |
| 9 | Qwen 3 1.7B | 80.8 | #1186 |
| 10 | Qwen 3 8B | 78.8 | #667 |
| 11 | Qwen 3 4B | 78.8 | #823 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=lingxidiagbench-dynamic-free-form-2-class-accuracy · How It Works · Data refreshed daily, snapshot 2026-10-11.