Medical Agent Task Completion on MedAgentBench
73.3AccuracyLLM-as-a-Verifier
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLM-as-a-Verifier2026.07 | 73.3 | — | — | |
| Opus 4.8Baseline Rank=#12026.07 | 70.2 | — | — | |
| G3.5 FlashBaseline Rank=#22026.07 | 66.3 | — | — | |
| GPT-5.5Baseline Rank=#32026.07 | 65.1 | — | — | |
| Candidate Pool2026.07 | — | 70.2 | 75 |