Protein Language Modeling Evaluation on ProteinLM Bench 944
77AccuracyGemini 3.1 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 3.1 ProType=Commercial LLMs, Setting=BioMedArena2026.05 | 77 | |
| Claude Sonnet 4.5Type=Commercial LLMs, Setting=BioMedArena2026.05 | 72.8 | |
| Gemini 3 FlashType=Commercial LLMs, Setting=BioMedArena2026.05 | 72.1 | |
| Claude Opus 4.6Type=Commercial LLMs, Setting=BioMedArena2026.05 | 71.8 | |
| SOTASource=[37]2026.05 | 71.5 | |
| Gemini 3.1 ProType=Commercial LLMs, Setting=Baseline2026.05 | 70.3 | |
| Claude Opus 4.5Type=Commercial LLMs, Setting=BioMedArena2026.05 | 69.8 | |
| Gemini 3 FlashType=Commercial LLMs, Setting=Baseline2026.05 | 65.7 | |
| Claude Sonnet 4.6Type=Commercial LLMs, Setting=BioMedArena2026.05 | 64.7 | |
| Claude Opus 4.6Type=Commercial LLMs, Setting=Baseline2026.05 | 64.4 | |
| Claude Opus 4.5Type=Commercial LLMs, Setting=Baseline2026.05 | 63.8 | |
| Claude Sonnet 4.6Type=Commercial LLMs, Setting=Baseline2026.05 | 60.3 | |
| Claude Sonnet 4.5Type=Commercial LLMs, Setting=Baseline2026.05 | 58.7 |