# Small Model as Master Orchestrator: Learning Unified Agent-Tool Orchestration with Parallel Subtask Decomposition

arXiv: https://arxiv.org/abs/2604.17009 · HTML: https://mlsanity.com/paper/2604.17009/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 40 references that are in the map.
- First posted on arXiv 18 Apr 2026.

## Builds on (most cited first, 27 of 27)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code](https://mlsanity.com/paper/2403.07974/) (2024)
- [MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark](https://mlsanity.com/paper/2406.01574/) (2024)
- [MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework](https://mlsanity.com/paper/2308.00352/) (2023)
- [Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning](https://mlsanity.com/paper/2503.09516/) (2025)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [Text Embeddings by Weakly-Supervised Contrastive Pre-training](https://mlsanity.com/paper/2212.03533/) (2022)
- [RouteLLM: Learning to Route LLMs with Preference Data](https://mlsanity.com/paper/2406.18665/) (2024)
- [A Survey on Large Language Models for Code Generation](https://mlsanity.com/paper/2406.00515/) (2024)
- [AFlow: Automating Agentic Workflow Generation](https://mlsanity.com/paper/2410.10762/) (2024)
- [Automated Design of Agentic Systems](https://mlsanity.com/paper/2408.08435/) (2024)
- [MathArena: Evaluating LLMs on Uncontaminated Math Competitions](https://mlsanity.com/paper/2505.23281/) (2025)
- [Multi-Step Reasoning with Large Language Models, a Survey](https://mlsanity.com/paper/2407.11511/) (2024)
- [Multi-Agent Collaboration via Evolving Orchestration](https://mlsanity.com/paper/2505.19591/) (2025)
- [AutoFlow: Automated Workflow Generation for Large Language Model Agents](https://mlsanity.com/paper/2407.12821/) (2024)
- [EvoFlow: Evolving Diverse Agentic Workflows On The Fly](https://mlsanity.com/paper/2502.07373/) (2025)
- [LLM-based Agents Suffer from Hallucinations: A Survey of Taxonomy, Methods, and Directions](https://mlsanity.com/paper/2509.18970/) (2025)
- [An Empirical Study on Reinforcement Learning for Reasoning-Search Interleaved LLM Agents](https://mlsanity.com/paper/2505.15117/) (2025)
- [Multi-Agent Collaboration via Cross-Team Orchestration](https://mlsanity.com/paper/2406.08979/) (2024)
- [DynTaskMAS: A Dynamic Task Graph-driven Framework for Asynchronous and Parallel LLM-based Multi-Agent Systems](https://mlsanity.com/paper/2503.07675/) (2025)
- [SwarmAgentic: Towards Fully Automated Agentic System Generation via Swarm Intelligence](https://mlsanity.com/paper/2506.15672/) (2025)
- [Dynamic Generation of Multi-LLM Agents Communication Topologies with Graph Diffusion Models](https://mlsanity.com/paper/2510.07799/) (2025)
- [Topological Structure Learning Should Be A Research Priority for LLM-Based Multi-Agent Systems](https://mlsanity.com/paper/2505.22467/) (2025)

## Built on by (most cited first, 2 of 2)

- [Reinforcement Learning for LLM-based Multi-Agent Systems through Orchestration Traces](https://mlsanity.com/paper/2605.02801/) (2026)
- [Visual Parallel Search: Learning to Search High-Resolution Images with Parallel Tile Inspection and Adaptive Zoom](https://mlsanity.com/paper/2609.37002/) (2026)

## Abstract (arXiv)

Multi-agent systems (MAS) demonstrate clear advantages in tackling complex problems by coordinating diverse agents and external tools. However, most existing orchestration methods rely on static workflows or serial agent scheduling, and are further constrained by heterogeneous interface protocols between tools and agents. This leads to high system complexity and poor extensibility. To mitigate these issues, we propose Agent-as-Tool, a unified parallel orchestration paradigm that abstracts both agents and tools into a standardized, learnable action space with protocol normalization and explicit state feedback. Building on this paradigm, we train a lightweight orchestrator, ParaManager, which decouples planning decisions from subtask solving, enabling state-aware parallel subtask decomposition, delegation, and asynchronous execution. For training, we adopt a two-stage ParaManager training pipeline. It improves robustness by incorporating supervised fine-tuning (SFT) trajectories equipped with recovery mechanisms, and further applies reinforcement learning (RL) to achieve an optimal balance among task success, protocol compliance, diversity, and reasoning efficiency. Experiments show that ParaManager achieves strong performance across multiple benchmarks and exhibits robust generalization under unseen model pools.