OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectories

TL;DR

OpenSeeker-v2 demonstrates that a simple supervised fine-tuning approach, fueled by informative high-difficulty trajectories and data synthesis, can achieve state-of-the-art search agent performance with minimal data and no heavy reinforcement learning.

Contribution

The paper introduces a novel data synthesis method and shows that a straightforward SFT approach can outperform complex pipelines in training frontier search agents.

Findings

OpenSeeker-v2 achieves state-of-the-art results on four benchmarks.

Training on only 10.6k data points suffices for top performance.

Purely academic development with no reinforcement learning is feasible.

Abstract

Deep search capabilities have become an indispensable competency for frontier Large Language Model (LLM) agents, yet their development remains dominated by industrial giants. The typical industry recipe involves a highly resource-intensive pipeline spanning pre-training, continual pre-training (CPT), supervised fine-tuning (SFT), and reinforcement learning (RL). In this report, we show that when fueled with informative and high-difficulty trajectories, a simple SFT approach could be surprisingly powerful for training frontier search agents. By introducing three simple data synthesis modifications: scaling knowledge graph size for richer exploration, expanding the tool set size for broader functionality, and strict low-step filtering, we establish a stronger baseline. Trained on merely 10.6k data points, our OpenSeeker-v2 achieves state-of-the-art performance across 4 benchmarks (30B-sized agents with ReAct paradigm): 46.0% on BrowseComp, 58.1% on BrowseComp-ZH, 34.6% on Humanity's Last Exam, and 78.0% on xbench, surpassing even Tongyi DeepResearch trained with heavy CPT+SFT+RL pipeline, which achieves 43.4%, 46.7%, 32.9%, and 75.0%, respectively. Notably, OpenSeeker-v2 represents the first state-of-the-art search agent within its model scale and paradigm to be developed by a purely academic team using only SFT. We are excited to open-source the OpenSeeker-v2 model weights and share our simple yet effective findings to make frontier search agent research more accessible to the community.