Paper 2605.04036
OpenSeeker-v2: Pushing the Limits of Search Agents with Informative and High-Difficulty Trajectories
- Published
- May 2026
- Research lab
- Independent
- Citations
- 6
- GitHub
- 766 stars
01 In brief
Summary
OpenSeeker-v2 is a search agent developed by an academic team at Shanghai Jiao Tong University, demonstrating that a simple supervised fine-tuning (SFT) approach can rival resource-intensive industrial pipelines when trained on high-quality, high-difficulty trajectories.
The authors introduce three data synthesis modifications: scaling the knowledge graph size for richer exploration, expanding the tool set for broader functionality, and applying strict low-step filtering to remove simple queries.
Trained on only 10.6k data points, OpenSeeker-v2 achieves state-of-the-art performance among 30B-sized ReAct agents on four benchmarks: 46.0% on BrowseComp, 58.1% on BrowseComp-ZH, 34.6% on Humanity's Last Exam, and 78.0% on xbench.
This surpasses Tongyi DeepResearch (43.4%, 46.7%, 32.9%, 75.0%) and RedSearcher, both trained with heavy CPT+SFT+RL pipelines.
The model also outperforms larger models like DeepSeek-V3.1-671B and GLM-4.6-357B.
The training data has an average of 64.67 tool calls per trajectory, higher than prior agents, indicating greater difficulty.
The authors open-source the model weights and code, making frontier search agent research more accessible.
02 From the paper
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.