Scientific Question Answering on HLE Drug Discovery
21.42Exact Match AccuracyMozi
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MoziModel Backend=Deepseek-V3.22026.03 | 21.42 | 6 | |
| BiomniModel Backend=Claude-4.5-Sonnet2026.03 | 17.86 | 5 | |
| MoziModel Backend=Qwen3-235B2026.03 | 17.86 | 5 | |
| STELLAModel Backend=Qwen3-30B2026.03 | 14.29 | 4 | |
| MiroThinker2026.03 | 14.29 | 4 | |
| MoziModel Backend=Qwen3-30B2026.03 | 14.29 | 4 | |
| Gemini-2.5-ProModel Backend=Gemini-2.5-Pro, Web Search Usage=true2026.03 | 10.71 | 3 | |
| BiomniModel Backend=Qwen3-30B2026.03 | 10.71 | 3 | |
| SciMaster2026.03 | 10.71 | 3 |