mlsanityData to 14 Oct 2026

Deep Research Agents: A Systematic Examination And Roadmap

arXiv 2506.18096 ↗First posted 22 Jun 2025cs.AIcited by 128

On the map: LLM agents and tool use → Web, search, and research agents

Abstract (arXiv)

The rapid progress of Large Language Models (LLMs) has given rise to a new category of autonomous AI systems, referred to as Deep Research (DR) agents. These agents are designed to tackle complex, multi-turn informational research tasks by leveraging a combination of dynamic reasoning, adaptive long-horizon planning, multi-hop information retrieval, iterative tool use, and the generation of structured analytical reports. In this paper, we conduct a detailed analysis of the foundational technologies and architectural components that constitute Deep Research agents. We begin by reviewing information acquisition strategies, contrasting API-based retrieval methods with browser-based exploration. We then examine modular tool-use frameworks, including code execution, multimodal input processing, and the integration of Model Context Protocols (MCPs) to support extensibility and ecosystem development. To systematize existing approaches, we propose a taxonomy that differentiates between static and dynamic workflows, and we classify agent architectures based on planning strategies and agent composition, including single-agent and multi-agent configurations. We also provide a critical evaluation of current benchmarks, highlighting key limitations such as restricted access to external knowledge, sequential execution inefficiencies, and misalignment between evaluation metrics and the practical objectives of DR agents. Finally, we outline open challenges and promising directions for future research. A curated and continuously updated repository of DR agent research is available at: {https://github.com/ai-agents-2030/awesome-deep-research-agent}.

Key facts. Cited by 128 papers in the map (data through 14 Oct 2026). Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use. Builds on 91 of its 131 references that are in the map. First posted on arXiv 22 Jun 2025.

Builds on

91 of 131 references.

Closest ancestors · 15

Same direction, newest first

  1. 2025WebDancer: Towards Autonomous Information Seeking Agency28 May
  2. 2025MLGym: A New Framework and Benchmark for Advancing AI Research Agents20 Feb
  3. 2025Improving Retrieval-Augmented Generation through Multi-Agent Reinforcement Learning25 Jan
  4. 2025Agentic Retrieval-Augmented Generation: A Survey on Agentic RAG15 Jan
  5. 2025Search-o1: Agentic Search-Enhanced Large Reasoning Models9 Jan
  6. 2025Agent Laboratory: Using LLM Agents as Research Assistants8 Jan
  7. 2024RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts22 Nov
  8. 2024ScienceAgentBench: Toward Rigorous Assessment of Language Agents for Data-Driven Scientific Discovery7 Oct
  9. 2024AssistantBench: Can Web Agents Solve Realistic and Time-Consuming Tasks?22 Jul
  10. 2024PersonaRAG: Enhancing Retrieval-Augmented Generation Systems with User-Centric Agents12 Jul
  11. 2024DS-Agent: Automated Data Science by Empowering Large Language Models with Case-Based Reasoning27 Feb
  12. 2024CoSearchAgent: A Lightweight Collaborative Search Agent with Large Language Models9 Feb
3 more
  1. 2023ReST meets ReAct: Self-Improvement for Multi-Step Reasoning LLM Agent15 Dec
  2. 2023MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation5 Oct
  3. 2023WebArena: A Realistic Web Environment for Building Autonomous Agents25 Jul

Other references · 76

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  6. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  7. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  8. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  9. 2025DAPO: An Open-Source LLM Reinforcement Learning System at Scalecited by 2,810
  10. 2023Retrieval-Augmented Generation for Large Language Models: A Surveycited by 2,212
  11. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  12. 2021WebGPT: Browser-assisted question-answering with human feedbackcited by 1,745
38 more
  1. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,444
  2. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,282
  3. 2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,195
  4. 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,128
  5. 2023GAIA: a benchmark for General AI Assistantscited by 1,128
  6. 2024The AI Scientist: Towards Fully Automated Open-Ended Scientific Discoverycited by 1,071
  7. 2022Atlas: Few-shot Learning with Retrieval Augmented Language Modelscited by 952
  8. 2025AlphaEvolve: A coding agent for scientific and algorithmic discoverycited by 867
  9. 2019ELI5: Long Form Question Answeringcited by 575
  10. 2023Active Retrieval Augmented Generationcited by 516
  11. 2025Humanity's Last Examcited by 491
  12. 2021A Dataset of Information-Seeking Questions and Answers Anchored in Research Paperscited by 486
  13. 2024A Survey on RAG Meeting LLMs: Towards Retrieval-Augmented Large Language Modelscited by 452
  14. 2025Accelerating scientific discovery with Co-Scientistcited by 401
  15. 2024LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitaskscited by 379
  16. 2024Measuring short-form factuality in large language modelscited by 373
  17. 2025R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learningcited by 336
  18. 2025WebThinker: Empowering Large Reasoning Models with Deep Research Capabilitycited by 335
  19. 2024Magentic-One: A Generalist Multi-Agent System for Solving Complex Taskscited by 232
  20. 2024Tool Learning with Large Language Models: A Surveycited by 223
  21. 2023OctoPack: Instruction Tuning Code Large Language Modelscited by 189
  22. 2024Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaborationcited by 189
  23. 2025WebWalker: Benchmarking LLMs in Web Traversalcited by 175
  24. 2024Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Modelscited by 154
  25. 2024CycleResearcher: Improving Automated Research via Automated Reviewcited by 148
  26. 2025OWL: Optimized Workforce Learning for General Multi-Agent Assistance in Real-World Task Automationcited by 131
  27. 2025Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolutioncited by 116
  28. 2024LongAlign: A Recipe for Long Context Alignment of Large Language Modelscited by 106
  29. 2025DeepReview: Improving LLM-based Paper Review with Human-like Deep Thinking Processcited by 100
  30. 2024SPA-Bench: A Comprehensive Benchmark for SmartPhone Agent Evaluationcited by 81
  31. 2024Beyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Modelscited by 76
  32. 2024AvaTaR: Optimizing LLM Agents for Tool Usage via Contrastive Reasoningcited by 72
  33. 2025AgentRxiv: Towards Collaborative Autonomous Researchcited by 72
  34. 2025Tool-Star: Empowering LLM-Brained Multi-Tool Reasoner via Reinforcement Learningcited by 72
  35. 2023ML-Bench: Evaluating Large Language Models and Agents for Machine Learning Tasks on Repository-Level Codecited by 65
  36. 2023TeleQnA: A Benchmark Dataset to Assess Large Language Models Telecommunications Knowledgecited by 63
  37. 2025SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesiscited by 55
  38. 2025AutoAgent: A Fully-Automated and Zero-Code Framework for LLM Agentscited by 52

50 most cited of 76 All 76 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 40; 2026: 88.

128 papers.

Sort:

By area:

  1. 2025Memento: Fine-tuning LLM Agents without Fine-tuning LLMscited by 127
  2. 2025ResearchRubrics: A Benchmark of Prompts and Rubrics For Evaluating Deep Research Agentscited by 84
  3. 2025Deep Research: A Survey of Autonomous Research Agentscited by 66
  4. 2025A Survey of LLM-based Deep Search Agents: Paradigm, Optimization, Evaluation, and Challengescited by 62
  5. 2025FutureX: An Advanced Live Benchmark for LLM Agents in Future Predictioncited by 59
  6. 2025Information Gain-based Policy Optimization: A Simple and Effective Approach for Multi-Turn Search Agentscited by 48
  7. 2025Agentic Web: Weaving the Next Web with AI Agentscited by 45
  8. 2026A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agentscited by 44
  9. 2025A Survey of Scientific Large Language Models: From Data Foundations to Agent Frontierscited by 37
  10. 2025Search Self-play: Pushing the Frontier of Agent Capability without Supervisioncited by 34
  11. 2025ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquirycited by 32
  12. 2025A Survey on Agentic Multimodal Large Language Modelscited by 23
38 more
  1. 2025COMPASS: Enhancing Agent Long-Horizon Reasoning with Evolving Contextcited by 20
  2. 2025Automatic Red Teaming LLM-based Agents with Model Context Protocol Toolscited by 18
  3. 2026Marco DeepResearch: Unlocking Efficient Deep Research Agents via Verification-Centric Designcited by 16
  4. 2025Reinforcement Learning Foundations for Deep Research Systems: A Surveycited by 14
  5. 2026DRACO: a Cross-Domain Benchmark for Deep Research Accuracy, Completeness, and Objectivitycited by 14
  6. 2025Co-Sight: Enhancing LLM-Based Agents via Conflict-Aware Meta-Verification and Trustworthy Reasoning with Structured Factscited by 13
  7. 2026OpenSkill: Open-World Self-Evolution for LLM Agentscited by 12
  8. 2025How Far Are We from Genuinely Useful Deep Research Agents?cited by 11
  9. 2026Adaptive Information Control for Search-Augmented LLM Reasoningcited by 11
  10. 2025VeriWeb: Verifiable Long-Chain Web Benchmark for Agentic Information-Seekingcited by 10
  11. 2025DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidencecited by 10
  12. 2025FlowSearch: Advancing deep research with dynamic structured knowledge flowcited by 10
  13. 2026Can We Predict Before Executing Machine Learning Agents?cited by 10
  14. 2026VSearcher: Long-Horizon Multimodal Search Agent via Reinforcement Learningcited by 9
  15. 2026How to Train Your Deep Research Agent? Prompt, Reward, and Policy Optimization in Search-R1cited by 8
  16. 2026DualSpec: Accelerating Deep Research Agents via Dual-Process Action Speculationcited by 8
  17. 2025Linguistic Nepotism: Trading-off Quality for Language Preference in Multilingual RAGcited by 7
  18. 2026SAGE: Benchmarking and Improving Retrieval for Deep Research Agentscited by 7
  19. 2026Memory Intelligence Agentcited by 7
  20. 2026AdaPlanBench: Evaluating Adaptive Planning in Large Language Model Agents under World and User Constraintscited by 7
  21. 2026Metis: Memory Foundation Modelcited by 7
  22. 2025K-Dense Analyst: Towards Fully Automated Scientific Analysiscited by 6
  23. 2025Erase to Improve: Erasable Reinforcement Learning for Search-Augmented LLMscited by 6
  24. 2025UFO3: Weaving the Digital Agent Galaxycited by 6
  25. 2026DLLM Agent: See Farther, Run Fastercited by 6
  26. 2026MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcomecited by 6
  27. 2025Network-Level Prompt and Trait Leakage in Local Research Agentscited by 5
  28. 2025SwarmSys: Decentralized Swarm-Inspired Agents for Scalable and Adaptive Reasoningcited by 5
  29. 2026Large Language Model Agent for User-friendly Chemical Process Simulationscited by 5
  30. 2026Why Your Deep Research Agent Fails? On Hallucination Evaluation in Full Research Trajectorycited by 5
  31. 2026TrinityGuard: A Unified Framework for Safeguarding Multi-Agent Systemscited by 5
  32. 2026Emergent Risks in Generative Multi-Agent Systemscited by 5
  33. 2026InfoSeeker: A Scalable Hierarchical Parallel Agent Framework for Web Information Seekingcited by 5
  34. 2026CausalGame: Benchmarking Causal Thinking of LLM Agents in Gamescited by 5
  35. 2026AREX: Towards a Recursively Self-Improving Agent for Deep Researchcited by 5
  36. 2025Fact2Fiction: Targeted Poisoning Attack to Agentic Fact-checking Systemcited by 4
  37. 2025Perspectra: Choosing Your Experts Enhances Critical Thinking in Multi-Agent Research Ideationcited by 4
  38. 2025LUT-LLM: Efficient Large Language Model Inference with Memory-based Computations on FPGAscited by 4

50 most cited of 128 All 128 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Deep Research of Deep Research: From Transformer to Agent, From AI to AI for Science
  2. 2025A Comprehensive Survey of Deep Research: Systems, Methodologies, and Applications
  3. 2025A Hierarchical Tree-based approach for creating Configurable and Static Deep Research Agent (Static-DRA)
  4. 2025From Web Search towards Agentic Deep Research: Incentivizing Search with Reasoning Agents
  5. 2026DR-Arena: an Automated Evaluation Framework for Deep Research Agents
  6. 2025DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents
  7. 2025ADRA-Bank: A Modular Benchmark for Academic Deep Research Agents
  8. 2025Deep Research Bench: Evaluating AI Web Research Agents