Chemical reasoning and prediction on ChemBench4K
93.66ProductGPT-5
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT-52026.05 | 93.66 | 78 | 43 | 99.33 | 64.62 | 75.72 | |
| SciCore-Mol2026.05 | 92 | 80.67 | 38.67 | 94.31 | 49.62 | 71.05 | |
| GPT-4o2026.05 | 91 | 58 | 45.66 | 96.32 | 60.25 | 70.25 | |
| Qwen3-8B-Chem2026.05 | 51 | 37.67 | 32.67 | 80.27 | 36.88 | 47.7 | |
| LlaSMol-Mistral-7B2026.05 | 22.33 | 23.33 | 32.33 | 35.79 | 21.88 | 27.13 | |
| Intern-S1-mini2026.05 | 22 | 24 | 30 | 22.41 | 22.25 | 24.13 | |
| Qwen3-8B2026.05 | 22 | 24 | 30 | 22.25 | 22.41 | 24.13 |