Find papers in one search
Stop hopping between platforms. Search arXiv, PubMed, OpenAlex and more from one place, and add any result straight to your library.
For students, PhDs and research teams
Distari summarises and organises your papers, helps you draft and cite from them, and brings your team into the same workspace.
Free forever · no card





01 / 05 · Library
Folders, smart filters and tags, with a summary, notes and references on every paper your team shares.
02 / 05 · Discover
New papers matched to your project brief and to the papers you already have, ready to add.
03 / 05 · Threads
Question one paper or the whole library, then compare methods and claims across papers side by side.
04 / 05 · Writing
Write in LaTeX, Typst or plain text, cite as you type and ask the library without leaving the draft.
05 / 05 · Overview
Goals, milestones and topic coverage for the team, with thin areas and suggested papers flagged.
Papers, notes and evidence sit together, so you always know where an idea came from.
One project carries each paper from the first search to the final citation.
Stop hopping between platforms. Search arXiv, PubMed, OpenAlex and more from one place, and add any result straight to your library.
Sort papers into folders, highlight and annotate as you read, and start from a summary to grasp each paper quickly.
Answers are grounded only in the papers you’ve collected, with citations that open the exact passage, across your whole library.
Write in LaTeX, Typst or plain text, with AI writing tools that already know your project’s papers.
Papers, highlights, notes and drafts stay in sync. Start on the web, an iPhone or iPad, the Mac, Windows or Linux today; Android is coming soon.
Share a project with your lab or co-authors. Everyone works from one library and sees each other’s annotations, while private notes stay private.
A walkthrough of Distari on real papers, start to finish.
The video couldn’t play here.
Open the tour video →Answers come only from the papers you’ve added. Every citation opens the passage it came from.
click [1] or [2] to see the source
Q Why did residual connections make very deep networks trainable?
Each block learns only a residual on top of an identity mapping, so extra depth starts from an easy default, and normalised inputs keep the signal stable.
“It is easier to optimize the residual mapping than to optimize the original, unreferenced mapping.”
“The distribution of each layer’s inputs changes during training.”
Start in any browser. Your papers, notes and drafts stay in sync.
A focused reader for long sessions, with offline notes. On macOS now, and as a beta on Windows and Linux.
Preview the desktop appRead, search and highlight papers on the go, with downloaded reading offline.
Preview the phone appA complete research library, free for good.
$0forever · no card
No time limit · your library never expires
$20per month
Renews monthly · cancel any time
$49instead of $60once · 3 months
You save $11 (18%) vs 3 months of monthly
Choose Pro in Plan & credits. Everything you add on Free carries over.
Prices in USD; US sales tax is added at checkout.