Computer Vision + Chemistry
Snap any chemical image — get every detail
Product label? Hand-drawn molecule? MSDS page? Claude Vision reads the image, extracts every component, and converts it to SMILES + measurable RDKit properties.
Extracted results
Upload an image to see extracted data here.
How it works: Your image is sent to the Claude Vision API which reads both text and molecular diagrams. Every detected ingredient is auto-resolved through PubChem for SMILES + InChIKey, then RDKit computes molecular weight, logP, and Lipinski violations. The output is structured, queryable data — not just transcribed text.