Knowledge base · no technical background needed
New here? This guide explains everything.
Never heard of “web scraping”? No problem. In five minutes you'll know what Joink does, why it's useful, and exactly how to use every feature.
What is “web scraping”, in plain words?
Imagine you're researching a topic. You open ten websites, read each one, and copy the useful bits into a document. It's slow, and a week later you can't remember which site a fact came from.
Web scraping is simply asking a computer to do that reading and copying for you. You give Joink the addresses (URLs) of public webpages. Joink visits each page the same way your browser does, and neatly files away what it finds:
And crucially, every saved fact remembers where it came from and when — the source link and timestamp travel with it forever. That's what makes Joink research you can actually trust and cite.
Who is Joink for?
🎓 Students
Collect sources for essays and projects with citations built in — no more "where did I read that?"
🔬 Researchers
Capture fast-changing web information in a structured, auditable format.
📣 Marketers
Track competitor pages, headlines and messaging across many sites at once.
💼 Job seekers
Save job postings and company pages before they change or disappear.
📊 Analysts
Turn scattered webpages into JSON/CSV you can drop into a spreadsheet or notebook.
✍️ Writers & journalists
Keep quotable, source-linked notes from every page you research.
How to use Joink, step by step
The whole journey — from a link to answered questions — looks like this:
1. Create a project
A project is just a folder for one piece of research — "Essay sources", "Competitor watch", anything. Click “New project” on your dashboard and give it a name.
2. Paste the web addresses
Copy the URL from your browser's address bar for each page you care about, and paste them in — one per line, up to 10 per run. Tick what you want captured (metadata, headings, text, links) and press “Extract content”.
3. Watch the extraction happen
Each address shows its own live status: queued → processing → completed. If one page fails (it happens — some sites block robots, some addresses are wrong), the others are still saved. Nothing is lost.
4. Explore your structured results
The results workspace has tabs: Overview (the big picture), Pages (full text), Headings (each page's outline), Links (everything the pages point to). Search by keyword, filter by page or heading level, copy anything, and open the original source in one click.
5. Ask questions in the Chat tab
Type a question like “What are these pages about?”. The answer comes only from your saved pages, with clickable citations showing exactly which page said it. If your pages don't contain the answer, Joink says so honestly instead of making something up.
6. Or just speak your question
Press the microphone, talk, and your words appear as editable text. Check the transcript, press send, and optionally have the answer read aloud. If your browser doesn't support voice, typing always works.
7. Export and come back anytime
Everything is saved automatically — close the tab and reopen it from your dashboard whenever. Need it elsewhere? Export to JSON or CSV (spreadsheet-ready); every row includes the source URL and timestamp.
Want to see it filled in before adding your own links? Press “Load demo project” on the dashboard — it creates a ready-made example with three extracted pages, one failed address, and a sample cited conversation.
The whole idea, in one picture
Frequently asked questions
I've never used a tool like this. Is it complicated?
No. If you can copy a link from your browser and paste it into a box, you can use Joink. The dashboard walks you through it, and the “Load demo project” button shows you a finished example before you start.
What exactly is a URL?
It's just a web address — the text in the bar at the top of your browser, like https://en.wikipedia.org/wiki/Chai. Copy it from any page you want Joink to read.
Is web scraping legal? Am I doing something shady?
Reading public webpages is what your browser does every day — Joink just organizes what's already publicly visible. It refuses private/internal addresses, honours sites' robots.txt wishes, never bypasses logins or paywalls, and rate-limits itself. You should still use the extracted content in line with each site's terms and your local laws, just as with manual copying.
What does “traceable” mean and why should I care?
Every fact Joink saves carries its source link, the exact time it was captured, and a confidence indicator. When someone asks “says who?”, you can answer with a click. Copy-pasting by hand loses all of that.
What can I do with the results?
Read and search them in the workspace, ask questions in plain language (typed or spoken) and get cited answers, copy anything to your clipboard, or export everything as JSON or CSV to open in Excel, Google Sheets, or your own tools.
What's the Chat tab? Is it like ChatGPT?
Similar to talk to, but with one big difference: it answers only from the pages you saved, and shows citations for every claim. If your saved pages don't contain the answer, it tells you that instead of guessing.
Why did one of my pages fail?
Common reasons: the address was mistyped, the site blocks automated readers (robots.txt), the page needs a login, or it's not a public webpage. A failed page never affects the others in the run — their results are still saved.
Why is some text marked “partial” or low confidence?
Some modern sites build their text with JavaScript after the page loads, so a simple read catches only part of it. Joink saves what it can, marks it honestly, and explains why — no silent gaps.
Do I have to pay?
No — the Free plan includes 3 projects and 5 page extractions a month, plus chat and voice questions. Pro (₹499/month) raises the limits and unlocks CSV export. Prices are set on the server and payments run through Razorpay.
Is my research private?
Yes. Your projects, results and conversations belong to your account only — enforced in the database itself (row-level security), not just in the interface. Voice questions are transcribed in your browser; raw audio is never uploaded or stored.
Joink only reads public pages, identifies itself honestly, respects sites' robots rules, and never breaks through logins, paywalls or CAPTCHAs. Good research starts with good manners.
Try it yourself — it's free