Stop searching. Start asking. Turn any document into an intelligent conversation partner.
Literature reviews, hypothesis generation, and finding connections across papers that you'd never spot manually. Advanced RAG strategies designed for scientific inquiry.
Connect Google Drive, Dropbox, OneDrive, Notion, Confluence, Slack or Zotero, search arXiv, PubMed and Semantic Scholar, or use your library from Claude Code via MCP. Upload PDFs, EPUBs, Word docs, spreadsheets, audio and video.
Shared workspaces, collaborative notes, and role-based access. Turn individual research into collective knowledge.
Your research data is always yours—export and migrate anytime, on our cloud or an Enterprise on-premise deployment.
Research is drowning in documents. Literature reviews take weeks. Finding connections between papers is manual detective work. Your team's collective knowledge stays siloed in individual folders.
The real challenge isn't finding information—it's understanding it. You need to:
Scrapalot transforms document chaos into structured knowledge. It's not just search—it's a research intelligence platform.
"I uploaded 50 research papers and asked it to compare methodologies across studies. Got a comprehensive answer in seconds with exact citations."
Perfect for literature reviews, hypothesis generation, grant writing, and discovering hidden connections across research domains. Import papers by DOI, arXiv or PubMed id, sync your Zotero library, and search academic databases from chat and Deep Research.
"Our PhD students complete literature reviews in days instead of months. The knowledge graph reveals relationships between papers we never noticed."
Accelerate research workflows, enable cross-departmental collaboration, and build institutional knowledge bases that persist beyond individual projects.
"We centralized 5 years of internal research reports. Now our scientists can query across all past experiments instantly."
Turn scattered research into searchable, queryable knowledge. Find prior art, avoid duplicating experiments, and build on past discoveries.
"Our technical documentation finally makes sense. New engineers onboard faster, and senior devs answer questions without context-switching."
Knowledge sharing without the bottleneck. Your documentation becomes a conversational partner, not a maze of markdown files.
Step 1: Upload & Connect Drag and drop PDFs, EPUBs, Word docs, spreadsheets, audio and video, or connect Google Drive, Dropbox, Notion, Zotero and more. Import papers straight from their DOI. We'll read everything—even scanned documents, and we transcribe recordings.
Step 2: Ask Anything
"What were the main findings in the Q3 reports?"
"Compare the pricing models mentioned in these contracts."
"Summarize all the security requirements across our documentation."
Step 3: Get Smart Answers
Not just keyword matches—actual understanding. With citations, so you can verify everything.
The Secret Sauce
We use 21 search strategies and 10 orchestrators, and let an AI agent automatically select the optimal approach for your question.
Three-Way Search Fusion:
Smart Context System:
Advanced Processing:
Knowledge Graph Enhancement:
You don't need to understand the internals—it just works. But for the curious, we've documented our architecture in detail.
Scrapalot is open-core: self-host the free, open-source Community Edition, or let us run the full product for you. See Editions for the full breakdown.
Run the core on your own infrastructure, no subscription.
GitHub: sime2408/scrapalot → · Self-host guide →
The managed cloud free tier — zero setup, perfect for individual researchers and students.
Apps: Use the web app in any browser, the native desktop app for Windows, macOS and Linux, or the native Android app.
The full AI research suite for one researcher.
Everything in Pro, scaled for research teams.
For large institutions and organizations.
Open-Core Model
Scrapalot is open-core. The core is free and open source; the advanced AI surfaces and the managed cloud are what fund development.
Community Edition (AGPL-3.0): Self-host the free, open-source core on your own infrastructure — no subscription, no quotas.
Researcher Plan: Free managed cloud tier for individual academics — in the web app, the desktop app or the Android app.
Pro / Team / Enterprise: The advanced AI features (deep research, knowledge graph, AI Scientist, voice, integrations) and team collaboration, fully managed on our cloud.
Your research data is always yours - export and migrate anytime.
Most tools do keyword matching. Scrapalot uses agentic RAG with 21 strategies and 10 orchestrators, tri-modal fusion (semantic + lexical + graph), and knowledge graphs. Ask "What methodologies do these papers share?" and get synthesis across sources with exact citations.
We built this because literature reviews were painful. Spending weeks reading papers, manually tracking citations, losing insights in scattered notes. If it frustrated us, we fixed it. The result: a research intelligence platform, not just document search.
Scrapalot is open-core: the core is free and open source, so you can self-host the Community Edition (AGPL-3.0) on your own infrastructure with Docker Compose. Prefer zero setup? The full product — advanced RAG, deep research, knowledge graphs, voice and team collaboration — is delivered as a managed cloud service with a free tier and paid subscriptions. Use Scrapalot in the web app (any browser), the native desktop app (Windows, macOS, Linux), or the native Android app. Individual academics get the core free on our cloud Researcher tier, or self-host it entirely for free.
Under the hood: Advanced search algorithms, intelligent routing, three-way fusion, context expansion, knowledge graphs, and 19 chunking strategies. On the surface: a simple chat interface. You get research-grade AI without needing a technical background.
Just Want to Try It?
Quick Start → - 10 minutes to your first answer
Want to Understand How It Works?
Architecture Deep Dive → - For the technically curious
Need Help?
FAQ → - Common questions
GitHub Discussions - Ask the community
Discord - Chat with us
Email Us - We actually respond