Year: 2026
Type: article
Abstract: Quantum chemistry is a foundational tool for chemistry, materials science, computational biology, and others. Despite its power, its application remains restricted to experts due to methodological complexity, software heterogeneity, and the need for informed interpretation. To bridge this accessibility gap, we introduce El Agente Quntur, a hierarchical, multi-agent system designed as a research collaborator for computational quantum chemistry. Quntur follows three strategies: (1) reasoning-driven decisions rather than hard-coded procedural policies, (2) general and composable actions, and (3) guided deep research integration of quantum-chemical reasoning with software logic and syntax. Instantiated in ORCA, Quntur reasons over software documentation and literature to plan, execute, adapt, and analyze in silico chemistry experiments. Benchmarking against Claude Code in the same software environment shows that Quntur's domain-specific agent architecture more faithfully translates open-ended chemical requests into defensible end-to-end computational protocols. We outline advances, bottlenecks, and a roadmap toward an autonomous end-to-end computational chemistry research agent.
Source: Chem
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $9,350
Funders: Schmidt Futures, U.S. Department of Energy, University of Minnesota, Merck KGaA, University of Toronto, National Research Foundation of Korea, Council for Science, Technology and Innovation, Canada First Research Excellence Fund, Defense Advanced Research Projects Agency, Advanced Research Projects Agency, Natural Sciences and Engineering Research Council of Canada, Natural Resources Canada, Defense Sciences Office, DARPA
OpenAlex ID: W7127984589