00358 收集到的 Deep Research 项目

大语言模型

发布日期: 2025-08-09

更新日期: 2025-08-09

前言

收集到的 Deep Research 项目。

Mermaid语法在线调试：https://mermaid.live

介绍

2025年2月2日OpenAI发布了Deep Research

src: https://openai.com/index/introducing-deep-research/

深度研究是OpenAI的下一个代理，可以独立为你工作——你给它一个提示，ChatGPT将找到、分析和综合数百个在线资源，以创建一个研究分析师级别的综合报告。由即将推出的针对网络浏览和数据分析进行优化的OpenAI o3模型的一个版本提供支持，它利用推理来搜索、解释和分析互联网上大量的文本、图像和PDF，根据需要对遇到的信息做出反应。

深度研究通过端到端强化学习在一系列领域的艰难浏览和推理任务中进行训练。通过这种训练，它学会了计划和执行多步骤轨迹来找到所需的数据，回溯并在必要时对实时信息做出反应。该模型还能够浏览用户上传的文件，使用python工具绘制和迭代图表，在其响应中嵌入生成的图表和来自网站的图像，并引用其来源中的特定句子或段落。作为这种培训的结果，它在许多专注于现实世界问题的公共评估中达到了新高。

DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents

src: https://huggingface.co/spaces/Ayanami0730/DeepResearch-Leaderboard

github: https://github.com/Ayanami0730/deep_research_bench

star: 258

Deep Research Agent From Scratch

github: https://github.com/swirl-ai/ai-angineers-handbook/tree/main/building_agents_from_scratch/deep_research_agent

star: 505

langchain-ai/open_deep_research

github: https://github.com/langchain-ai/open_deep_research

star: 7328

langchain-ai/local-deep-researcher

github: https://github.com/langchain-ai/local-deep-researcher

star: 7.9k

assafelovic/gpt-researcher

github: https://github.com/assafelovic/gpt-researcher

star: 22.9k

zilliztech/deep-searcher

github: https://github.com/zilliztech/deep-searcher

star: 6.8k

mshumer/OpenDeepResearcher

github: https://github.com/mshumer/OpenDeepResearcher

star: 2.7k

dzhng/deep-research

github: https://github.com/dzhng/deep-research

star: 17.4k

flowchart TB
    subgraph Input
        Q[User Query]
        B[Breadth Parameter]
        D[Depth Parameter]
    end

    DR[Deep Research] -->
    SQ[SERP Queries] -->
    PR[Process Results]

    subgraph Results[Results]
        direction TB
        NL((Learnings))
        ND((Directions))
    end

    PR --> NL
    PR --> ND

    DP{depth > 0?}

    RD["Next Direction:
    - Prior Goals
    - New Questions
    - Learnings"]

    MR[Markdown Report]

    %% Main Flow
    Q & B & D --> DR

    %% Results to Decision
    NL & ND --> DP

    %% Circular Flow
    DP -->|Yes| RD
    RD -->|New Context| DR

    %% Final Output
    DP -->|No| MR

    %% Styling
    classDef input fill:#7bed9f,stroke:#2ed573,color:black
    classDef process fill:#70a1ff,stroke:#1e90ff,color:black
    classDef recursive fill:#ffa502,stroke:#ff7f50,color:black
    classDef output fill:#ff4757,stroke:#ff6b81,color:black
    classDef results fill:#a8e6cf,stroke:#3b7a57,color:black

    class Q,B,D input
    class DR,SQ,PR process
    class DP,RD recursive
    class MR output
    class NL,ND results

jina-ai/node-DeepResearch

github: https://github.com/jina-ai/node-DeepResearch

star: 4.7k

---
config:
  theme: mc
  look: handDrawn
---
flowchart LR
 subgraph Loop["until budget exceed"]
    direction LR
        Search["Search"]
        Read["Read"]
        Reason["Reason"]
  end
    Query(["Query"]) --> Loop
    Search --> Read
    Read --> Reason
    Reason --> Search
    Loop --> Answer(["Answer"])

flowchart TD
    Start([Start]) --> Init[Initialize context & variables]
    Init --> CheckBudget{Token budget<br/>exceeded?}
    CheckBudget -->|No| GetQuestion[Get current question<br/>from gaps]
    CheckBudget -->|Yes| BeastMode[Enter Beast Mode]

    GetQuestion --> GenPrompt[Generate prompt]
    GenPrompt --> ModelGen[Generate response<br/>using Gemini]
    ModelGen --> ActionCheck{Check action<br/>type}

    ActionCheck -->|answer| AnswerCheck{Is original<br/>question?}
    AnswerCheck -->|Yes| EvalAnswer[Evaluate answer]
    EvalAnswer --> IsGoodAnswer{Is answer<br/>definitive?}
    IsGoodAnswer -->|Yes| HasRefs{Has<br/>references?}
    HasRefs -->|Yes| End([End])
    HasRefs -->|No| GetQuestion
    IsGoodAnswer -->|No| StoreBad[Store bad attempt<br/>Reset context]
    StoreBad --> GetQuestion

    AnswerCheck -->|No| StoreKnowledge[Store as intermediate<br/>knowledge]
    StoreKnowledge --> GetQuestion

    ActionCheck -->|reflect| ProcessQuestions[Process new<br/>sub-questions]
    ProcessQuestions --> DedupQuestions{New unique<br/>questions?}
    DedupQuestions -->|Yes| AddGaps[Add to gaps queue]
    DedupQuestions -->|No| DisableReflect[Disable reflect<br/>for next step]
    AddGaps --> GetQuestion
    DisableReflect --> GetQuestion

    ActionCheck -->|search| SearchQuery[Execute search]
    SearchQuery --> NewURLs{New URLs<br/>found?}
    NewURLs -->|Yes| StoreURLs[Store URLs for<br/>future visits]
    NewURLs -->|No| DisableSearch[Disable search<br/>for next step]
    StoreURLs --> GetQuestion
    DisableSearch --> GetQuestion

    ActionCheck -->|visit| VisitURLs[Visit URLs]
    VisitURLs --> NewContent{New content<br/>found?}
    NewContent -->|Yes| StoreContent[Store content as<br/>knowledge]
    NewContent -->|No| DisableVisit[Disable visit<br/>for next step]
    StoreContent --> GetQuestion
    DisableVisit --> GetQuestion

    BeastMode --> FinalAnswer[Generate final answer] --> End

u14app/deep-research

github: https://github.com/u14app/deep-research

star: 3.8k

flowchart TB
    A[Research Topic]:::start

    subgraph Propose[Propose your ideas]
        B1[System asks questions]:::process
        B2[System outputs the research plan]:::process
        B3[System generates SERP queries]:::process
        B1 --> B2
        B2 --> B3
    end

    subgraph Collect[Information collection]
        C1[Initial research]:::collection
        C1a[Retrieve local research resources based on SERP queries]:::collection
        C1b[Collect information from the Internet based on SERP queries]:::collection
        C2[In-depth research]:::recursive
        Refine{More in-depth research needed?}:::decision

        C1 --> C1a
        C1 --> C1b
        C1a --> C2
        C1b --> C2
        C2 --> Refine
        Refine -->|Yes| C2
    end

    Report[Generate Final Report]:::output

    A --> Propose
    B3 --> C1

    %% Connect the exit from the loop/subgraph to the final report
    Refine -->|No| Report

    %% Styling
    classDef start fill:#7bed9f,stroke:#2ed573,color:black
    classDef process fill:#70a1ff,stroke:#1e90ff,color:black
    classDef recursive fill:#ffa502,stroke:#ff7f50,color:black
    classDef output fill:#ff4757,stroke:#ff6b81,color:black
    classDef collection fill:#a8e6cf,stroke:#3b7a57,color:black
    classDef decision fill:#c8d6e5,stroke:#8395a7,color:black

    class A start
    class B1,B2,B3 process
    class C1,C1a,C1b collection
    class C2 recursive
    class Refine decision
    class Report output

nickscamara/open-deep-research

github: https://github.com/nickscamara/open-deep-research

star: 5.9k

LearningCircuit/local-deep-research

github: https://github.com/LearningCircuit/local-deep-research

star: 3.3k

btahir/open-deep-research

github: https://github.com/btahir/open-deep-research

star: 2k

qx-labs/agents-deep-research

github: https://github.com/qx-labs/agents-deep-research

star: 573

IterativeResearcher Flow

flowchart LR
    A["User Input<br>- query<br>- max_iterations<br>- max_time<br>- output_instructions"] --> B

    subgraph "Deep Research Loop"
        B["Knowledge<br>Gap Agent"] -->|"Current gaps<br>& objective"| C["Tool Selector<br>Agent"]
        C -->|"Tool queries<br>(run in parallel)"| D["Tool Agents<br>- Web Search<br>- Crawler<br>- Custom tools"]
        D -->|"New findings"| E["Observations<br>Agent"]
        E --> |"Thoughts on findings<br>and research strategy"| B
    end

    E --> F["Writer Agent<br>(final output<br>with references)"]

DeepResearcher Flow

flowchart LR
    A["User Input<br>- query<br>- max_iterations<br>- max_time"] --> B["Planner Agent"]
    
    B -->|"Report plan<br>(sections & background context)"| D2
    
    subgraph Parallel["Parallel Section Research"]
        D1["IterativeResearcher<br>(Section 1)"]
        D2["IterativeResearcher<br>(Section 2)"]
        D3["IterativeResearcher<br>(Section 3)"]
    end
    
    D1 -->|"Section 1<br>Draft"| E["Proofreader<br>Agent"]
    D2 -->|"Section 2<br>Draft"| E
    D3 -->|"Section 3<br>Draft"| E
    
    E --> F["Final<br>Research<br>Report"]

结语

第三百五十八篇博文写完，开心！！！！

今天，也是充满希望的一天。

LuYF-Lemon-love

https://luyf-lemon-love.space/2025/08/09/00358-shou-ji-dao-de-deep-research-xiang-mu/

本博客所有文章除特別声明外，均采用 CC BY 4.0 许可协议。转载请注明来源 LuYF-Lemon-love !

大语言模型

本篇

00358 收集到的 Deep Research 项目

2025-08-09 大语言模型

大语言模型

00357 🧪 Python `unittest` 单元测试框架教程

2025-08-07 Python

Python

前言

介绍

DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents

Deep Research Agent From Scratch

langchain-ai/open_deep_research

langchain-ai/local-deep-researcher

assafelovic/gpt-researcher

zilliztech/deep-searcher

mshumer/OpenDeepResearcher

dzhng/deep-research

jina-ai/node-DeepResearch

u14app/deep-research

nickscamara/open-deep-research

LearningCircuit/local-deep-research

btahir/open-deep-research

qx-labs/agents-deep-research

IterativeResearcher Flow

DeepResearcher Flow

结语

谢谢小主！