Search-o1: Agentic Search-Enhanced Large Reasoning Models

By Xiaoxi Li, Guanting Dong, Jiajie Jin, Yuyao Zhang, Yujia Zhou, Yutao Zhu, Peitian Zhang, Zhicheng Dou. Conference on Empirical Methods in Natural Language Processing, 2025

446 citations recorded by the source corpus.

Large reasoning models (LRMs) like OpenAI-o1 have demonstrated impressive long stepwise reasoning capabilities through large-scale reinforcement learning. However, their extended reasoning processes often suffer from knowledge insufficiency, leading to frequent uncertainties and potential errors. To address this limitation, we introduce \textbf{Search-o1}, a framework that enhances LRMs with an agentic retrieval-augmented generation (RAG) mechanism and a Reason-in-Documents module for refining retrieved documents. Search-o1 integrates an agentic search workflow into the reasoning process, enabling dynamic retrieval of external knowledge when LRMs encounter uncertain knowledge points. Additionally, due to the verbose nature of retrieved documents, we design a separate Reason-in-Documents module to deeply analyze the retrieved information before injecting it into the reasoning chain, minimizing noise and preserving coherent reasoning flow. Extensive experiments on complex reasoning tasks in science, mathematics, and coding, as well as six open-domain QA benchmarks, demonstrate the strong performance of Search-o1. This approach enhances the trustworthiness and applicability of LRMs in complex reasoning tasks, paving the way for more reliable and versatile intelligent systems. The code is available at \url{https://github.com/sunnynexus/Search-o1}.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/5b86fbaf12fd50710066e3995e33ece2185f8ada

Citations in the corpus

• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning (2025, unknown): “(2) Inference with Retrieval : Retrieval-Augmented Generation (RAG) (Lewis et al., 2020), IRCoT (Trivedi et al., 2022a), and Search-o1 (Li et al., 2025).”
• Large Language Models for Information Retrieval: A Survey (2023, unknown): “Despite LLMs being capable of identifying potential user intents of the query [220], they lack awareness of the resulting retrieval quality of the rewritten query.”
• WebThinker: Empowering Large Reasoning Models with Deep Research Capability (2025, unknown): “For example, Search-o1 [26] incorporates an agentic RAG framework and a Reason-in-Documents module to merge retrieval with reasoning via prompt engineering.”
• MG2-RAG: Multi-Granularity Graph for Multimodal Retrieval-Augmented Generation (2026, unknown): “Hierarchical query decomposition [40,76], iterative retrieval-reasoning loops [22,66], and agentic workflows [41,45,69] now enable models to emulate structured research processes.”
• ReasonIR: Training Retrievers for Reasoning Tasks (2025, unknown): “However, these systems like R1-Searcher (Li et al., 2025) rely on retrieval models such as BGE (Xiao et al., 2024b), which perform well on semantic retrieval benchmarks but struggle with reasoning-heavy tasks (Su et al., 2024).”
• A Survey of Model Architectures in Information Retrieval (2025, unknown): “The commonly used smoothing methods (Zhai and Lafferty, 2004; Zhai et al., 2008) include Jalinek-Mercer smoothing, Dirichlet smoothing, etc.”