extracurricular-ai

extracurricular-ai / open-deep-research-with-web-ui

Public

🔍 AI-powered deep research agent with Web UI. Built on smolagents, featuring process-based architecture, MetaSo/DuckDuckGo search, and Docker deployment. Apache 2.0 licensed.

ai-agent docker flask gunicorn litellm
16
1
100% credibility
Found Mar 15, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A self-hosted web app replicating OpenAI's Deep Research agent with parallel background tasks, multi-model support, and session history.

How It Works

1
🔍 Discover Open Deep Research

You hear about this free tool that lets AI do deep internet research for you, just like a smart assistant.

2
🚀 Get it running easily

Download and start it with simple instructions, no complicated setup needed.

3
🔗 Connect your AI brain

Link it to your favorite AI service so it can think and search smartly.

4
💭 Ask your big question

Type in a tough research question and hit run — watch the magic unfold live.

5
📱 Run many at once

Fire off several questions in parallel and switch between them anytime.

6
⏳ Check back anytime

Close your browser, come back later — your researches keep going safely.

✅ Get your deep insights

Review detailed answers with sources, ready to use for work or curiosity.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is open-deep-research-with-web-ui?

This Python project delivers an AI-powered deep research agent with a sleek web UI, replicating OpenAI's Deep Research for tackling complex queries via web search and multi-agent reasoning. Users fire off questions like "EU AI Act compliance checklist" and get structured reports with sources, supporting models from OpenAI, Claude, DeepSeek, or local Ollama via LiteLLM. It handles vision tasks like PDF analysis or YouTube transcripts, with Docker for one-command deploys and CLI fallback.

Why is it gaining traction?

Parallel background research stands out—queue multiple long-running tasks (10-30 mins each) without blocking the UI, persist sessions in SQLite for later replay, even after browser closes. Free DuckDuckGo search works out-of-box (no API key), plus fallbacks like MetaSo; no Node build needed for the Preact frontend. Hits 55% on GAIA benchmark, beating many open alternatives in multi-step reasoning.

Who should use this?

Data scientists batch-analyzing trends, product managers researching competitors, or indie devs prototyping AI-powered deep search tools. Ideal for teams self-hosting on VMs/Raspberry Pi needing concurrent queries without babysitting tabs. Skip if you need production-scale beyond Gunicorn workers.

Verdict

Promising for AI-powered deep research workflows, especially with DeepSeek integration and background mode—grab it via Docker for quick experiments. Low 1.0% credibility from 16 stars signals early maturity; solid docs but watch for edge cases in media tools.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.