Text-based Molecule Design on ChEBI-20-MM
38Exact MatchChemAmp (Stage 1+2)
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| ChemAmp (Stage 1+2)Model Category=Agent with Tool Amplification, Evaluation Protocol=0-shot, Stage=1+22025.05 | 38 | 93 | 8.68 | 96 | 92 | 87 | 80 | 86 | |
| ChemCrowModel Category=Agent with Tool Orchestration, Evaluation Protocol=0-shot, Tool Set Size=182025.05 | 34 | 77 | 45.4 | 68 | 84 | 80 | 76 | 80 | |
| ChemAmp (Stage 1)Model Category=Agent with Tool Amplification, Evaluation Protocol=0-shot, Stage=12025.05 | 34 | 87 | 12.63 | 94 | 85 | 80 | 74 | 80 | |
| Text+Chem T5Model Category=Chemistry specialized models, Evaluation Protocol=0-shot2025.05 | 32 | 85 | 16.87 | 94 | 90 | 82 | 75 | 82 | |
| ChemDFM-13BModel Category=Chemistry specialized models, Evaluation Protocol=0-shot2025.05 | 32 | 85 | 11.58 | 94 | 81 | 73 | 67 | 74 | |
| ChemToolAgentModel Category=Agent with Tool Orchestration, Evaluation Protocol=0-shot, Tool Set Size=292025.05 | 28 | — | — | 90 | — | — | — | 77 | |
| MolXPTModel Category=Chemistry specialized models, Evaluation Protocol=0-shot2025.05 | 22 | — | — | 98 | 86 | 76 | 67 | 76 | |
| Llama3-70bModel Category=Generalist LLMs, Evaluation Protocol=0-shot2025.05 | 3 | 57 | 46.63 | 78 | 57 | 40 | 30 | 42 | |
| Mol-InstructionModel Category=Chemistry specialized models, Evaluation Protocol=0-shot2025.05 | 2 | 35 | 41.4 | 100 | 41 | 23 | 15 | 26 | |
| Deepseek-R1Model Category=Generalist LLMs, Evaluation Protocol=0-shot2025.05 | 2 | 56 | 92.29 | 57 | 48 | 38 | 31 | 39 | |
| GPT-4oModel Category=Generalist LLMs, Evaluation Protocol=0-shot2025.05 | 1 | 57 | 52.85 | 91 | 71 | 54 | 38 | 54 |