QUASAR: A Universal Autonomous System for Atomistic Simulation and a Benchmark of Its Capabilities

TL;DR

QUASAR presents a production-ready universal autonomous system for atomistic simulation that orchestrates DFT, ML potentials, MD, and MC workflows through a three-agent framework (Strategist, Operator, Evaluator). It combines double-pass planning, scalable memory management, and hybrid knowledge retrieval within containerized, HPC-optimized environments to enable autonomous scientific discovery with minimal human intervention. Across a three-tier benchmark, QUASAR demonstrates progression from reliable single tasks to autonomous frontier screening, including NiO band-gap challenges addressed via auto-improvement and frontier tasks aligned with published results. The work argues that reasoning-driven orchestration can approach expert-level autonomy in computational chemistry, while acknowledging the need for broader benchmarks and continued development of evaluation standards and human–AI collaboration paradigms.

Abstract

The integration of large language models (LLMs) into materials science offers a transformative opportunity to streamline computational workflows, yet current agentic systems remain constrained by rigid tool-calling approaches and narrowly scoped agents. In this work, we introduce QUASAR, a universal autonomous system for atomistic simulation designed to facilitate production-grade scientific discovery. QUASAR autonomously orchestrates complex multi-scale workflows across diverse methods, including density functional theory, machine learning potentials, molecular dynamics, and Monte Carlo simulations. The system incorporates robust mechanisms for adaptive planning, context-efficient memory management, and hybrid knowledge retrieval to navigate real-world research scenarios without human intervention. We benchmark QUASAR against a series of three-tiered tasks, progressing from routine tasks to frontier research challenges such as photocatalyst screening and novel material assessment. These results suggest that QUASAR can function as a general atomistic reasoning system rather than a task-specific automation framework. They also provide initial evidence supporting the potential deployment of agentic AI as a component of computational chemistry research workflows, while identifying areas requiring further development.

Figures (2)

• Figure 1: Overview of the QUASAR architecture, where dashed lines represent optional feedback paths for iterative refinement.
• Figure 2: A three-stage pipeline illustrating how the Operator coordinates end-to-end simulation tasks.