Work

Can AI Help Chemists To Solve NMR?

Year: 2026

Type: preprint

Abstract: Chemists read 1 H and 13 C NMR daily, and it is tempting to paste the peak lists into a large language model. Given only those two peak lists for 105 molecules balanced by molecular complexity and compound classes, six frontier LLMs (Claude Opus 4.8, DeepSeek V4 Pro, Gemini 3.1 Pro, GPT-5.5, Grok 4.3, Qwen3.7-Max) and four specialized solvers (BLIND, NMRMind, NMRPeak, NMR-Solver) were compared. The best LLM (Gemini 3.1 Pro) solved 24% of structures, the best specialized model (BLIND) solved 69% of them. All models drift toward smaller molecules and drop groups silent in ¹H NMR like halogens, nitriles, diazo.

Source: ChemRxiv

DOI: 10.26434/chemrxiv.15006195/v2

Authors: Fedor Kliuev, Ivan Smirnov, Mikhail A. Losev, Oleg I. Afanasyev, Denis Chusov

Institutions: Russian Academy of Sciences, A. N. Nesmeyanov Institute of Organoelement Compounds

Language: English

Topic: Molecular spectroscopy and chirality

Subfield: Spectroscopy

Field: Chemistry

Domain: Physical Sciences

Open Access status: green

Funder: Ministry of Science and Higher Education of the Russian Federation

OpenAlex ID: W7213981463

Cites: 105

Cited by:

Related to: 10