O-Researcher: An Open Ended Deep Research Model via Multi-Agent Distillation and Agentic RL

2026-01-07

arXiv.org

AI Summary

This work addresses the performance gap between open-source and closed-source large language models, which stems largely from the scarcity of high-quality training data for the former. To bridge this gap, the authors propose an end-to-end multi-agent distillation framework that leverages collaborative agents to simulate complex tool-integrated reasoning processes, thereby automatically generating research-grade instruction data. The framework employs a two-stage training strategy combining supervised fine-tuning and agent-oriented reinforcement learning, without relying on any proprietary datasets. This approach significantly enhances the reasoning capabilities of open-source models on deep research tasks, achieving new state-of-the-art results among open-source models across multiple established benchmarks and effectively narrowing the performance disparity with closed-source counterparts.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Multiagent Systems: Multiagent Learning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

The performance gap between closed-source and open-source large language models (LLMs) is largely attributed to disparities in access to high-quality training data. To bridge this gap, we introduce a novel framework for the automated synthesis of sophisticated, research-grade instructional data. Our approach centers on a multi-agent workflow where collaborative AI agents simulate complex tool-integrated reasoning to generate diverse and high-fidelity data end-to-end. Leveraging this synthesized data, we develop a two-stage training strategy that integrates supervised fine-tuning with a novel reinforcement learning method, designed to maximize model alignment and capability. Extensive experiments demonstrate that our framework empowers open-source models across multiple scales, enabling them to achieve new state-of-the-art performance on the major deep research benchmark. This work provides a scalable and effective pathway for advancing open-source LLMs without relying on proprietary data or models.

Problem

• large language models
• open-source
• training data
• performance gap
• instructional data

Innovation

• multi-agent distillation
• agentic reinforcement learning
• synthetic instructional data
• open-source LLMs
• tool-integrated reasoning