New Anthropic Science Blog: Making Claude a chemist. To manipulate a molecule, chemists first need to understand its structure. Their main tool is NMR spectroscopy. We found Opus 4.7 matches—and on some tasks beats—dedicated NMR software. Read more: https://t.co/1jUvz7wdhV
Anthropic's Claude Opus 4.7 Matches, Outperforms Dedicated NMR Software for Chemists
Anthropic· Updated
Anthropic's Claude Opus 4.7 model now performs at or above the level of specialized NMR spectroscopy software for chemistry tasks. This demonstrates a general-purpose AI's ability to handle precise analytical work in a complex scientific domain, moving beyond general conversational AI.
- Models evaluated
- Opus 4.7, Opus 4.6, Sonnet 4.6
- Compared to
- ChemDraw, MestReNova
- Inverse prediction compounds
- 15
- Opus 4.7 1H error
- +/-0.079 ppm
- Opus 4.7 13C error
- +/-1.37 ppm
This performance is notable because Claude Opus 4.7 is a general-purpose model, not specifically fine-tuned for chemistry. It suggests that frontier models can now handle highly specialized scientific tasks that traditionally required bespoke software. The ability to work in reverse, proposing structures from NMR data, is a task existing software often leaves to human chemists.
For routine data prediction, Claude Opus 4.7 performed comparably or better than tools like ChemDraw and MestReNova. Its ability to elucidate structures from 1D NMR data and molecular formulas, without specialized 2D NMR or training, makes complex analysis more accessible. Anthropic plans to extend Claude's helpfulness in areas like reading chemical structures, reaction reasoning, and understanding chemical literature.
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →





