Comparison
CurTask vs Connected Papers: a graph of papers, then a review you can write
Checked on 11 October 2026
This page is for students and researchers who use Connected Papers to see the shape of a field and want a place to do the reading, comparing and writing that comes after. Connected Papers is known for building a visual graph around one paper, so you can spot related work, earlier foundations and later studies at a glance. CurTask is the workspace for the next step: one project where your papers, your analysis and your chapter stay together.
At a glance
| Capability | CurTask | Connected Papers |
|---|---|---|
| Find related papers | Yes — catalogue search and "papers like these" | Yes — its main focus |
| Keep and read your PDFs, with highlights and notes | Yes — one Global Library and a reader | Not a focus |
| Scanned PDFs read in English and Malay | Yes | Not a focus |
| AI analysis of each paper | Yes — every value with a page and verbatim quote | Not a focus |
| Papers found without a PDF | Saved and marked "No PDF"; add the PDF later | Not a focus |
| Literature matrix (papers × attributes) | Yes — AI fills columns, you accept each value | Separate tool needed |
| Qualitative coding of papers | Yes — by hand on every plan; AI on paid plans | Separate tool needed |
| Themes and agree/contradict links | Yes — synthesis map, links you mark yourself | Separate tool needed |
| Download all your data | Yes — one ZIP in open formats, any time | Paper lists as BibTeX |
| Writing with citations | Yes — 8 styles, linked back to the PDF page | Separate tool needed |
| Interface in Bahasa Melayu | Yes — English and Bahasa Melayu | No — English only |
| Pricing | Free plan; Student and Researcher plans in ringgit | Free: 5 graphs a month; Academic and Business plans |
| Where your work is stored | Servers in Malaysia | Microsoft Azure; region not stated |
From a map of the field to the papers themselves
A graph tells you which papers belong together. The review asks a harder question: what do those papers actually say, and how do they compare? CurTask is built for that question. Your papers live in one Global Library, each stored once, with a reader for highlights and notes. Scanned PDFs, including Malay ones, are read so their text can be searched and analysed, and files are made smaller with every word of their text kept.
CurTask also helps you keep finding papers while you work. You can search open research catalogues from inside the app, add open-access PDFs in one click, and ask for papers like the ones already in your project. A paper with no open-access PDF is still saved with its details and marked "No PDF"; add the PDF later from the paper's drawer. The AI analyses only papers that have their PDF, so nothing is charged for the others.
A literature matrix with evidence in every cell
In CurTask you describe a matrix column in plain words, such as "research design" or "key finding", see the credit cost first, and run it across your papers. Every AI value keeps the page number and a verbatim quote, and CurTask checks that the quote is really on that page. If no source is found, the cell says "No source found" instead of guessing. Values stay marked Suggested until you accept, edit or restore them, and the matrix can be exported as a spreadsheet (CSV).
What CurTask learns about a paper travels with it. If the same paper appears in your next project, its analyses come along, and reusing them costs no credits.
Coding, themes and writing in the same project
When a review needs deeper reading, you can code passages in your papers and see them as coloured strips beside the text, write memos on them, and group codes into themes on the synthesis map. You can mark which papers agree, contradict or extend each other, then send a theme to your draft with its passages. Coding by hand is on every plan; on paid plans the AI codes every paper of a project at the push of a button and sorts your codes into categories and themes, each suggestion yours to accept or dismiss.
The Writing lens is a modular editor with binder, corkboard and outline views. Cite with @ in one of eight styles, including APA, Harvard and IEEE; the bibliography rebuilds when you change style. Each citation can link back to its passage in the PDF, shown as the evidence line, and optional marks show sentences that have no citation yet. Compile to Word, PDF, Markdown or HTML. Beside the draft, the writing sidebar builds on the analysis you have accepted (matrix values, coded passages, themes and links): up to three cards per paragraph, contradictions first, each with its page and verified quote.
Local by design
CurTask's interface is in English and Bahasa Melayu. Your papers, notes and writing are stored on servers in Malaysia, visible only to you, and never used to train AI. Plans are priced in ringgit, with a free plan and a Student plan for university students, and every AI action shows its credit cost before it runs. You can download all your data at any time, on every plan, as one ZIP file in open formats, with your references as BibTeX, RIS and CSL-JSON.
When Connected Papers might suit you
If you are at the very start of a topic and want a quick visual sense of which papers cluster together and which are the classics, Connected Papers is a simple and well-regarded way to get it, and you can try it for free.
Using both
The two fit together well. Build a graph in Connected Papers to spot the papers you need, download their PDFs, and drag them into CurTask. CurTask fills in the details from each paper's DOI, which you can edit, and checks for duplicates so nothing is stored or analysed twice. From there, the matrix, coding and writing all happen in one project.
Map the field wherever you like, then bring your papers into CurTask and start on the free plan.
Try CurTask on your own papers
Import a few PDFs, fill a literature matrix and see the page and quote behind every value.
Free plan with monthly AI credits. No card needed.
Product names are trademarks of their owners. Information as of 11 October 2026, from public sources. If something here is out of date, tell us at hello@hackedu.tech.