Replaces traditional quantitative structure-activity relationship models with fine-tuned GPT models. Enables zero-shot and few-shot learning for molecular property estimation.
Tech Stack
Jupyter NotebookChemistry & Materials
Why gptchem?
• Leverages pre-trained knowledge
• Simple API for fine-tuning
• Supports multiple LLM backends
Limitations
• API costs for closed models
• Potential for chemical hallucinations
• Black-box decision making
2/17/2026
Last Update
43
Forks
23
Issues
MIT
License
Financial Leak Detected
Stop the "SaaS Tax"
Your team could be burning cash. Switching to gptchem instantly boosts your runway.
