Large language model (LLM) multi-agent systems typically rely on rigid orchestration, committing either to flat per-query routing or to hand-engineered task decomposition, so decomposition depth, worker choice, and inference budget are not jointly optimized under one objective. We introduce Uno-Orchestra, a unified orchestration policy that selectively decomposes a task and dispatches each subtask to an admissible (model, primitive) pair, with both decisions learned together from curated RL trajectories grounded in real worker interactions. Against 22 baselines on a 13-benchmark suite spanning math, code, knowledge, long-context, and agentic tool-use, Uno-Orchestra reaches 77.0% macro pass@1, roughly 16% above the strongest workflow baseline, at roughly an order of magnitude lower per-query cost, advancing the accuracy-efficiency frontier of selective delegation.
@article{arxiv.2605.05007,
title = {Uno-Orchestra: Parsimonious Agent Routing via Selective Delegation},
author = {Zhiqing Cui and Haotong Xie and Jiahao Yuan and Cheng Yang and Hanqing Wang and Yuxin Wu and Yifan Wu and Siru Zhong and Tao Yu and Yifu Guo and Siyu Zhang and Xinlei Yu and Qibing Ren and Usman Naseem},
journal= {arXiv preprint arXiv:2605.05007},
year = {2026}
}