Baichuan M3 235B: benchmark results
Provider: Baichuan. Released 2026-02-06. Access: Open.
Aggregate Rating 1665 ± 26, rank #525 of 10916 rated models, from 69 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LingxiDiagBench (Dynamic, APA-Guided) - 4-Class Accuracy | 34.5 | Accuracy (%, 0-100) on four-way classification (pure depress | 100 |
| LingxiDiagBench (Dynamic, Clinical, Free-Form) - 4-Class Accuracy | 43.5 | Accuracy (%, 0-100) on four-way classification (pure depress | 100 |
| Medmarks - HealthBench | 65.64 | Score (%) | 100 |
| Medmarks - HealthBench Consensus | 67.33 | Score (%) | 100 |
| Medmarks - HealthBench Hard | 48.21 | Score (%) | 100 |
| Medmarks - MedDialog | 94.61 | Score (%) | 100 |
| Medmarks - MedicationQA | 92.74 | Score (%) | 100 |
| LingxiDiagBench (Dynamic, APA-Guided) - 2-Class Accuracy | 88.5 | Accuracy (%, 0-100) on binary depression-versus-anxiety clas | 95.8 |
| Medmarks - MedXpertQA Understanding | 39.05 | Score (%) | 88.6 |
| LingxiDiagBench (Dynamic, Symptom-Tree) - 4-Class Accuracy | 30.5 | Accuracy (%, 0-100) on four-way classification (pure depress | 87.5 |
| Medmarks - M-ARC | 68.33 | Score (%) | 87.5 |
| Medmarks - MedMCQA | 77.91 | Score (%) | 87.5 |
Interactive version: theaggregate.ai/model?slug=baichuan-m3-235b · How It Works · Data refreshed daily, snapshot 2026-11-08.