Long-form QA on MultiMed-X YO
4.45Overall ScoreMED-COREASONER
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MED-COREASONERbackbone=GPT-5.12026.01 | 4.45 | 4.46 | 4.79 | 4.7 | 4.23 | 91 | |
| GPT-5.12026.01 | 4.09 | 4.16 | 4.31 | 4.595 | 4.08 | 81 | |
| GPT-5.22026.01 | 4.065 | 4.135 | 4.225 | 4.55 | 4.06 | 79.5 | |
| Claude-3.5-haiku2026.01 | 3.545 | 3.555 | 3.65 | 4.435 | 3.585 | 60 | |
| GPT-4o2026.01 | 3.29 | 3.325 | 3.4 | 4.095 | 3.205 | 45.5 |