Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 28 min ago
PreprintCC BY-NC-NDChemRxiv

Can AI Help Chemists To Solve NMR?

  • Fedor Kliuev1
  • Ivan Smirnov1
  • Mikhail A. Losev1
  • Oleg I. Afanasyev1
  • Denis Chusov1
  1. 1A.N. Nesmeyanov Institute of Organoelement Compounds of Russian Academy of Sciences, 28 Vavilova St., Moscow, Russian Federation, 119234
Open at sourcePDFDOICite

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.

Related

  1. Plant-Based Proteins Have a Flavor Issue. Chemists Can Help with That
    Carolyn Wilke
    ACS Central SciencePeer-reviewedCC BYFull textGeneral chemistryDOAJ
  2. SMDD-Bench: Can LLMs Solve Real-World Small Molecule Drug Design Tasks?
    Kevin Han, Rui Zhang, Kathy Y. Wei +3
    arXiv (Cornell University)PreprintCC BYFull text6 citedMedicinal chemistry & drug discoveryOpenAlex
  3. How the fruit fly connectome can help autonomous materials discovery
    Kinston Ackölf, Taylor D. Sparks
    ChemRxivPreprintCC BYFull textMaterials & polymersChemRxiv

Cite

BibTeXRIS
BibTeX
@misc{kliuev2026help,
  title = {{Can AI Help Chemists To Solve NMR?}},
  author = {Fedor Kliuev and Ivan Smirnov and Mikhail A. Losev and Oleg I. Afanasyev and Denis Chusov},
  howpublished = {ChemRxiv},
  year = {2026},
  month = sep,
  doi = {10.26434/chemrxiv.15006195/v2},
  url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15006195/v2},
  note = {CC BY-NC-ND},
}
RIS
TY  - UNPB
TI  - Can AI Help Chemists To Solve NMR?
AU  - Fedor Kliuev
AU  - Ivan Smirnov
AU  - Mikhail A. Losev
AU  - Oleg I. Afanasyev
AU  - Denis Chusov
PY  - 2026
DA  - 2026/09/22
PB  - ChemRxiv
DO  - 10.26434/chemrxiv.15006195/v2
UR  - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15006195/v2
AB  - 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.
ER  -