Deep Research

by U14 AppAppResearch something

Turns a research question into a written report with sources in a few minutes

View on GitHub

The problem it solves

You need a solid overview of a topic by tomorrow, and all you have is twenty open browser tabs and no notes.

For Students and researchersPeople preparing reports

Before: You search, open tab after tab and copy findings into a document by hand.
After: You type a topic and answer a few questions; it searches, then writes a report with citation links you can edit and extend.

Stars: ★ 4.7k
License: MIT
Cost: Usage-based
Platforms: Web
Runs: Cloud · Self-hosted

Screenshots

From the project's own repository.

The Deep Research web app with four steps: research topic, questions, information collection and final report — Image: U14 App (MIT). Original file

Overview

Deep Research is an open-source web app from U14 App that writes a research report on a topic you give it. It asks follow-up questions, searches the web through your chosen AI model or a search service, and puts the findings into a report you can edit, translate and turn into a knowledge graph. You can use the hosted version in your browser with your own API key, or deploy your own copy to Vercel, Cloudflare or Docker; research history stays in your browser. It suits students, analysts and anyone who needs a quick first overview of a topic.

It works with many model providers, including a free Gemini API key, and keeps your research history in your own browser rather than on a server.

Try these first

Start with the first one: it is a small task you can check easily.

Write a short report on the history of [your hometown], with sources.First try · uses API credits
Research the pros and cons of heat pumps for an older house in a cold climate.
Make the final report shorter and easier to read, then translate it into Spanish.
Use the PDF I added as a local resource and compare it with what you find online.sends file to model

Example prompts written by Agent Flows Hub.

Features

From the creator's documentation.

Feature	What it does	Runs
Follow-up questions	Asks you questions about the topic before it starts collecting information	Your AI provider
Web search	Uses the model's own search or services such as Tavily, Exa or SearXNG	Search service
Report editor	Edit the report, change its reading level or length, or translate it	Your AI provider
Knowledge graph	Turns a finished report into a knowledge graph in one click	Your AI provider
Local files	Add text, Office or PDF files as a local knowledge base for the research	Your browser
History	Keeps past research in your browser so you can reopen it and research again	Your browser

Good to know

Give it a topic, answer its clarifying questions, and let it search and write a structured report with citation links. You can then edit the report, change its reading level or length, and ask for more research on one part.

Limits

Reports are written by the model you connect and can contain mistakes or weak sources. Open the cited links before you rely on a claim.
Your question goes to the AI provider and search service you configure. If you deploy your own copy with server-side keys, set an access password so others can't use them.

Notes by Agent Flows Hub (original project name: deep-research), written from the creator's documentation, not from running it ourselves. Deep Research README: getting started and deployment

Stuck setting it up?

Tell us what happened. We take a look first; if it needs paid work, you get an estimate before anything starts.

Get help

How to start

Install itPaste the install request into an AI agent you already use, or run the official command yourself or follow the official guide.macOS, Windows, Linuxdocker pull xiangfa/deep-research:latest docker run -d --name deep-research -p 3333:3000 xiangfa/deep-research
Open itFollow the first-run steps in the app.
Give it a first taskAsk for a short report on a topic you already know well, such as the history of your hometown, and check three of its cited sources.

Steps follow the creator's instructions. The request tells your agent to read them first and wait for your OK.

Before you start

A computer with a web browser.
An API key from an AI provider, for example a free Gemini API key from Google AI Studio
Optional: a search service key such as Tavily or Exa, for models that can't search
For your own copy: a Vercel or Cloudflare account, or Docker 20 or newer
Pay per use. The AI service it calls charges for what you use.

Details

Made by: U14 App
License: MIT Terms
Platforms: Web
Runs on: Cloud, Self-hosted server

View source Official docs Report a problem

For AI agents

This page as a Markdown install brief:

/en/library/u14app-deep-research/install.md

Other apps

DeepChat Brings many AI models, skills, connectors and agents into one desktop chat app
DeepTutor Turns your notes and documents into a personal tutor that explains, quizzes and plans
Deta Surf Keeps web pages, PDFs and videos in local notebooks and writes cited notes from them
Dia Answers questions from your open tabs and work apps, and drafts briefs, reports and decks
DocsGPT Turns your files and sites into an AI assistant that answers with its sources
Dograh Builds AI voice agents that answer or make phone calls, using a visual flow editor

See all 113 apps

GitHub