Qubrio: High-Performance Quantum Compilation via Multi-Agent LLM Collaboration
Abstract
Large Language Models (LLMs) offer an adaptable alternative to traditional quantum compilation heuristics, which bottleneck progress by demanding costly manual redesigns whenever hardware evolves. However, naive single-pass LLM approaches fail under the unique challenges introduced by quantum compilation. We introduce Qubrio, a practical agentic framework that significantly reduces heuristic redesigns through a three-level decomposition. By (1) partitioning programs into sequential operation stages, (2) deploying specialized agents for placement, routing and optimization, and (3) using disaggregated feedback for precise error correction, our system reliably navigates complex physical constraints. Beyond serving as a direct compiler, our LLM framework also uncovers novel strategies that facilitate convoy-style shuttling, substantially enhancing hardware concurrency without sacrificing fidelity. We further integrate these strategies into the previous state-of-the-art (SOTA) compiler, recovering a significant fraction of the LLM's performance advantages. On realistic workloads, our compiler achieves significant improvements in hardware runtime and program fidelity over the SOTA baseline PowerMove, seamlessly adapting to new hardware capabilities. Qubrio features an open user interface for practical use, and its code is publicly available at https://github.com/crosaa/Qubrio.
Related
- QuPort: Topology-, Port-, and Congestion-Aware Compilation for Modular Multi-QPU Quantum Systems
- Compiling Together: High-Throughput Distributed Quantum Computing via Multi-Compilation
- Quantum circuit compilation with constant overhead
- pfoqcompiler: Efficient Compilation of Quantum Control in Recursive First-Order Quantum Programs
- A Compilation Framework for Quantum Simulation of Non-unitary Dynamics
BibTeX
@misc{ruan2026qubrio,
title = {{Qubrio: High-Performance Quantum Compilation via Multi-Agent LLM Collaboration}},
author = {Jixuan Ruan and Zhuo Cui and Zhengding Hu and Zhongkai Yu and Xiang Fang and Yue Guan and Jason Ludmir and Phillip E. Weinberg and Xiuzhe Luo and Shengtao Wang and Sonia Lopez Alarcon and Hezi Zhang and Yufei Ding},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.02269},
eprint = {2611.02269},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.02269},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Qubrio: High-Performance Quantum Compilation via Multi-Agent LLM Collaboration
AU - Jixuan Ruan
AU - Zhuo Cui
AU - Zhengding Hu
AU - Zhongkai Yu
AU - Xiang Fang
AU - Yue Guan
AU - Jason Ludmir
AU - Phillip E. Weinberg
AU - Xiuzhe Luo
AU - Shengtao Wang
AU - Sonia Lopez Alarcon
AU - Hezi Zhang
AU - Yufei Ding
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.02269
UR - https://arxiv.org/abs/2611.02269
AB - Large Language Models (LLMs) offer an adaptable alternative to traditional quantum compilation heuristics, which bottleneck progress by demanding costly manual redesigns whenever hardware evolves. However, naive single-pass LLM approaches fail under the unique challenges introduced by quantum compilation. We introduce Qubrio, a practical agentic framework that significantly reduces heuristic redesigns through a three-level decomposition. By (1) partitioning programs into sequential operation stages, (2) deploying specialized agents for placement, routing and optimization, and (3) using disaggregated feedback for precise error correction, our system reliably navigates complex physical constraints. Beyond serving as a direct compiler, our LLM framework also uncovers novel strategies that facilitate convoy-style shuttling, substantially enhancing hardware concurrency without sacrificing fidelity. We further integrate these strategies into the previous state-of-the-art (SOTA) compiler, recovering a significant fraction of the LLM's performance advantages. On realistic workloads, our compiler achieves significant improvements in hardware runtime and program fidelity over the SOTA baseline PowerMove, seamlessly adapting to new hardware capabilities. Qubrio features an open user interface for practical use, and its code is publicly available at https://github.com/crosaa/Qubrio.
ER -