Skip to content
Agent Search Engine.

System class · No. 07

Research & Data Agents

Agents that search, analyze, and synthesize information, documents, and data.

40 records · independent index · how we rank

Curated products

Alphabetical — stars don't measure commercial adoption
Adobe Acrobat AI AssistantPlatformAsk cited document questions and organize shared PDF Spaces.Paid Acrobat AI plan/add-on or eligible bundleLicense not checkedRecord checkedAI for DatabasePlatformAsk database questions, build dashboards and configure data-driven workflows.FreemiumLicense not checkedRecord checkedAskMedPlatformResearch medical questions with publisher-described source references.License not checkedRecord checkedConsensusAgentAI search engine over 200M+ scientific papers.FreemiumLicense not checkedRecord checkedDesearchInfrastructureRetrieve web and social search results through API, SDK and MCP access.Paid productLicense not checkedRecord checkedDoclingInfrastructureConvert complex documents into structured data with local or configured processing pipelines.MIT code; hardware, model and optional service costsOpen sourcerecorded license: MITRecord checked · Recent push 71kGitHub starsElicitAgentResearch assistant for paper search, PDF summaries and literature reviews.Costs need recheckingLicense not checkedRecord checkedFlashCrafter Growth BriefPlatformGenerate a private local-business growth report with visible evidence and assumptions.FreemiumLicense not checkedRecord checkedGemini Deep ResearchAgentReview a research plan, then generate a report from selected sources.FreemiumLicense not checkedRecord checkedGemini NotebookPlatformStudy your sources with cited answers and generated learning material.Free Standard with limits; paid upgradesLicense not checkedRecord checkedHarveyPlatformDomain-specific AI agents for legal and professional services work.Paid productLicense not checkedRecord checkedKhanmigoPlatformUse guided tutoring and teaching tools with role-specific eligibility.Eligible teacher tools free; learner access paidLicense not checkedRecord checkedNolynAgentResearch ecommerce markets and products using supported marketplace and social data.Paid productLicense not checkedRecord checkedPerplexityAgentAI answer engine with deep research agents that cite their sources.Free search; paid feature and credit eligibility variesLicense not checkedRecord checkedScribizPlatformExtract speech and on-screen text from video through web and developer tools.FreemiumLicense not checkedRecord checkedTemu Products Scraper (Zentra)InfrastructureFetch Temu catalogue rows through an Apify Actor for data workflows.Paid productLicense not checkedRecord checkedUnstructured TransformPlatformExtract document content through a hosted browser workflow or API.Starter page allowance; then paid usageLicense not checkedRecord checked

Repository-listed tools

Stars measure repository interest. A public repository does not establish an open-source license, product quality or active users; check each record and its official documentation.

→ Compare records marked open source in research & data agents

By language: Python research & data agents

Order this page by
firecrawlInfrastructureThe API to search, scrape, and interact with the web at scale.Costs need recheckingOpen sourceTypeScript · recorded license: AGPL-3.0Record checked · Recent push 202kGitHub starsScraplingInfrastructureAn adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!Costs not checkedOpen sourcePython · recorded license: BSD-3-ClauseRecord checked · Recent push 104kGitHub starsTrendRadarAgentAI-driven public opinion & trend monitor with multi-platform aggregation, RSS, and smart alerts.Costs not checkedOpen sourcePython · recorded license: GPL-3.0Record checked · Recent push 64kGitHub starsBettaFishAgentA multi-agent public-opinion analysis assistant anyone can use -- reconstructs the real shape of public sentiment and forecasts where it is heading. Built from scratch, no framework dependencies.Costs not checkedOpen sourcePython · recorded license: GPL-2.0Record checked · Recent push 43kGitHub starskhojAgentYour AI second brain. Self-hostable. Get answers from the web or your docs. Build custom agents, schedule automations, do deep research.Costs not checkedOpen sourcePython · recorded license: AGPL-3.0Record checked 38kGitHub starsWeKnoraPlatformOpen-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.Costs not checkedLicense not checkedGoRecord checked · Recent push 38kGitHub starsstormAgentAn LLM-powered knowledge curation system that researches a topic and generates a full-length report with citations.Costs not checkedOpen sourcePython · recorded license: MITRecord checked 32kGitHub starsgpt-researcherAgentAn autonomous agent that conducts deep research on any data using any LLM providersCosts not checkedOpen sourcePython · recorded license: Apache-2.0Record checked · Recent push 31kGitHub starsFastGPTPlatformFastGPT is a knowledge-based platform built on the LLMs, offers a comprehensive suite of out-of-the-box capabilities such as data processing, RAG retrieval, and visual AI workflow orchestration,Costs not checkedLicense not checkedTypeScriptRecord checked · Recent push 30kGitHub starshaystackFrameworkOpen-source AI orchestration framework for building context-engineered, production-ready LLM applications.Costs not checkedOpen sourcePython · recorded license: Apache-2.0Record checked · Recent push 27kGitHub starsDeepResearchAgentTongyi Deep Research, the Leading Open-source Deep Research AgentCosts not checkedOpen sourcePython · recorded license: Apache-2.0Record checked 20kGitHub starslocal-deep-researchAgent~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google,...). 10+ search engines - arXiv, PubMed, your private documents.Costs not checkedOpen sourcePython · recorded license: MITRecord checked · Recent push 9.6kGitHub stars
For agent buildersMake your product part of the discovery.Explore advertising
MiroThinkerAgentMiroThinker is a deep research agent optimized for complex research and prediction tasks. Our latest models, MiroThinker-1.7, achieves 74.0 and 75.3 on the BrowseComp and BrowseComp Zh, respectively.Costs not checkedOpen sourcePython · recorded license: Apache-2.0Record checked 8.5kGitHub starsdeep-searcherAgentOpen Source Deep Research Alternative to Reason and Search on Private Data. Written in Python.Costs not checkedOpen sourcePython · recorded license: Apache-2.0Record checked 8.4kGitHub starsunstractPlatformLLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline WorkflowsCosts not checkedOpen sourcePython · recorded license: AGPL-3.0Record checked · Recent push 7.5kGitHub starsDeepAnalyzeAgentDeepAnalyze is the first agentic LLM for autonomous data science.Costs not checkedOpen sourcePython · recorded license: MITRecord checked · Recent push 4.8kGitHub starsms-agentFrameworkMS-Agent: a lightweight framework to empower agentic execution of complex tasksCosts not checkedOpen sourcePython · recorded license: Apache-2.0Record checked · Recent push 4.4kGitHub starsSidekickAgentA native macOS app that allows users to chat with a local LLM that can respond with information from files, folders and websites on your Mac without installing any other software.Costs not checkedOpen sourceSwift · recorded license: MITRecord checked 3.3kGitHub starsMiroFlowFrameworkTop-1 on 5+ benchmarks | Web UI | Supports MiroThinker, Claude, Kimi, OpenAICosts not checkedOpen sourcePython · recorded license: Apache-2.0Record checked 3.2kGitHub starsToolUniverseSDK / libraryDemocratizing AI scientists with ToolUniverseCosts not checkedOpen sourcePython · recorded license: Apache-2.0Record checked · Recent push 1.8kGitHub starsAuto-Deep-ResearchAgent"Your Fully-Automated Personal AI Assistant"Costs not checkedLicense not checkedPythonRecord checked 1.8kGitHub starsAdalaFrameworkAdala: Autonomous DAta (Labeling) Agent frameworkCosts not checkedOpen sourcePython · recorded license: Apache-2.0Record checked · Recent push 1.6kGitHub starsTrustRAGFrameworkTrustRAG:The RAG Framework within Reliable input,Trusted outputCosts not checkedLicense not checkedPythonRecord checked 1.3kGitHub stars
How to choose research & data agents

Research and data agents search, read, and synthesize information — across the open web, academic sources, or your own documents — into answers you can act on, ideally with citations. The decision comes down to source quality and coverage, how transparently the agent cites what it used, and whether it can work over your private or internal data as well as the public web. For anything high-stakes, verifiable citations you can click and check matter far more than fluent prose: the failure mode of these tools is confident synthesis of wrong or outdated sources, so treat them as accelerators, not authorities. Look at how the agent handles conflicting sources, whether it distinguishes primary from secondary material, and how it deals with recency on fast-moving topics.

If you're pointing it at your own corpus, evaluate the connectors, how it chunks and retrieves, and whether answers stay grounded in your documents rather than drifting to the model's memory. Pricing ranges from free open-source pipelines you run yourself to usage-based hosted products, and deeper research runs cost more in tokens and time. The most useful tools make their reasoning and sources inspectable so you can audit an answer instead of just trusting it — verify anything consequential before you rely on it, and prefer transparency over confidence.

Building one yourself? How to build a RAG knowledge assistant →

Frequently asked

What is a research agent?
A research agent gathers and synthesizes information across many sources — web, papers, or your documents — and produces summarized, ideally cited, answers.
How do I choose a research agent?
Weigh source coverage and quality, how transparently it cites, whether it can search your private data, and how it handles conflicting or uncertain information.
Can research agents be trusted?
Treat them as accelerators, not authorities: prefer tools that cite sources you can check, and verify anything consequential.