Baichuan M2 32B: benchmark results
Provider: Baichuan. Released 2025-09-02. Access: Open.
Aggregate Rating 1579 ± 19, rank #1258 of 10916 rated models, from 116 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LingxiDiagBench (Dynamic, Clinical, APA-Guided) - 2-Class Accuracy | 82.5 | Accuracy (%, 0-100) on binary depression-versus-anxiety clas | 100 |
| LingxiDiagBench (Dynamic, Clinical, APA-Guided) - 4-Class Accuracy | 43 | Accuracy (%, 0-100) on four-way classification (pure depress | 100 |
| LingxiDiagBench (Dynamic, Clinical, APA-Guided) - Consultation Quality | 3.98 | Consultation quality (1-6): mean of five Likert dimensions ( | 100 |
| LingxiDiagBench (Dynamic, Clinical, Symptom-Tree) - 4-Class Accuracy | 44 | Accuracy (%, 0-100) on four-way classification (pure depress | 100 |
| LingxiDiagBench (Static, Clinical) - 12-Class Accuracy | 24.7 | Accuracy (times 100, 0-100) on twelve-category ICD-10 diagno | 100 |
| MedMT-Bench (Text Subset) | 46.32 | Accuracy (%) on the 285 text-only dialogues of MedMT-Bench's | 100 |
| LingxiDiagBench (Dynamic, Symptom-Tree) - 4-Class Accuracy | 30.5 | Accuracy (%, 0-100) on four-way classification (pure depress | 87.5 |
| LingxiDiagBench (Dynamic, Clinical, Free-Form) - 12-Class Accuracy | 44 | Accuracy (%, 0-100) on twelve-category ICD-10 diagnosis (F20 | 83.3 |
| LingxiDiagBench (Dynamic, Clinical, Free-Form) - 2-Class Accuracy | 87.5 | Accuracy (%, 0-100) on binary depression-versus-anxiety clas | 83.3 |
| Medmarks - SCTpublic | 71.16 | Score (%) | 81.8 |
| BRIDGE Medical Leaderboard - CoT | 38.3 | Average Performance (%) | 80.2 |
| LingxiDiagBench (Dynamic, Clinical, APA-Guided + MRD-RAG) - 2-Class Accuracy | 80 | Accuracy (%, 0-100) on binary depression-versus-anxiety clas | 80 |
Interactive version: theaggregate.ai/model?slug=baichuan-m2-32b · How It Works · Data refreshed daily, snapshot 2026-11-08.