Skip to content
AI study companion for research papers

Turn any research paper into a course you can master.

Upload a paper and get illustrated chapter explanations, equations worked through step by step, runnable code, flashcards and audio lessons, plus a Teacher that answers from your own library.

$14 USD / month · Optional 7-day free trial · Cancel anytime

Fig. 1What is Structured Learning? A short tour of the app. Watch on YouTube
What you get

Everything a hard paper needs, in one place

Structured Learning reads the paper with you: it explains, derives, implements and quizzes, and every answer points back to the original text.

Illustration: the Research Desk with the chapter list on the left, a chapter explanation with a teaching chart in the middle, and citation pills linked to the paper’s own page text, figure and equation on the right.
Fig. 2The Research Desk: the explanation in the middle, the paper’s own evidence on the right.
Explanations

Chapter-by-chapter explanations you can check

Structured Learning reads the whole paper, plans a course, and explains it one chapter at a time in plain language, from the motivation to the core method and the results.

  • Optional teaching figures generated for every chapter
  • Citations link each passage to the page, figure, table or equation it came from
  • Continue reading picks up at the chapter where you stopped
Illustration: the attention equation expanded in five annotated steps, from the query-key product to the weighted sum of values.
Fig. 3Scaled dot-product attention, derived in five steps.
Equations

Equations, one step at a time

Open any equation for a guided walkthrough. It is expanded step by step, every symbol is defined, and each step says what changed and why.

  • Every symbol and dimension explained
  • Steps checked symbolically with SymPy where possible
  • Linked back to where the equation appears in the paper
Illustration: attention implemented in three incremental code steps, next to flashcards and a multiple-choice check.
Fig. 4Attention in three code steps, then a quick check.
Code

Code that builds up step by step

See the paper’s method implemented incrementally. Each step adds a few lines to the last one, so you understand the code instead of copying it.

  • Implementations that grow one step at a time
  • Flashcards and comprehension checks to test what you learned
Illustration: Teacher answering a question with numbered citations to seven papers in the library.
Fig. 5One question, answered from seven papers.
Teacher

A Teacher that knows your whole library

Ask about one paper or all of them. Teacher answers from your own library, cites the passages it used, and draws a diagram when words are not enough.

  • Answers grounded in your papers, with sources
  • Teaching diagrams on request
  • Ask by text or by voice
Illustration: a knowledge map linking concepts from biology, chemistry, physics, astronomy and AI, grown from a stack of research papers.
Fig. 6Biology, chemistry, physics, astronomy and AI on one map.
Knowledge map

One map of everything you read

Every paper you add joins a knowledge map that links concepts across papers and fields, so you can see how ideas connect and where a new paper fits.

  • Concepts linked across papers and fields
  • Explore it as an interactive graph
Illustration: a goal interview ("understand diffusion models") leading to a study plan of five concepts in order, each with its papers, a 10-question quiz on the current concept and a search for open-access papers on the last one.
Fig. 7One goal, five concepts in study order, each with its papers.
Learning plans

A study plan built around your goal

Tell a short interview what you want to learn. It checks what you already know, then lays out the concepts to study in order, each with the papers that teach it.

  • Concepts you already know are marked as known
  • Pass a 10-question quiz to finish each concept
  • Find open-access papers for a concept that has none

Also included

  • Flashcards & quizzesSpaced repetition and comprehension checks
  • Audio lessonsListen to any chapter, narrated
  • Peer reviewsPublic reviews from OpenReview, eLife, PLOS and more
  • DiscoveryFind papers on arXiv and OpenAlex
  • Reading modesFive reading themes, from warm ivory to dark
  • Agent accessLet your AI agents read your library over MCP
How it works

From paper to course in four steps

  1. 01

    Create an account and connect OpenRouter

    Sign up with email or Google, then paste your OpenRouter API key in Settings. It powers every AI feature, and you pick the models.

  2. 02

    Find or upload a paper

    Find papers on arXiv and OpenAlex with Discovery and add them straight to your library, or upload a PDF, Word or LaTeX file up to 50 MB. Choose text-only or illustrated explanations and watch the course take shape chapter by chapter.

  3. 03

    Build a learning plan

    A short interview asks about your goal and checks what you already know, then lays out the concepts to study in order, each with its papers. Generate their explanations, pass a 10-question quiz to finish a concept, or find open-access papers for one that has none.

  4. 04

    Study your way

    Read the explanations, work through equations and code, review flashcards, listen to audio lessons, or ask Teacher anything.

Inside the app, a tutorial and guided tours walk you through every feature.

Transparent AI costs

You bring the AI key. We never mark it up.

Structured Learning doesn’t resell AI. Every model call runs on your own OpenRouter account, so you choose the models and see exactly what each paper costs.

Paid to Structured Learning

$14 USD / month

The platform

  • Every feature on this page
  • Your private library: papers, explanations, chats and audio
  • New features and fixes as they ship
  • Support by email
Paid to OpenRouter

Pay as you go

AI usage

  • Billed by OpenRouter at each model’s price, from your credits
  • You pick the models for text, images and narration
  • Set a spending limit and see every cent on the Usage page
Pricing

One plan with everything included

No tiers, no feature gates and no long-term contract. Pay today, or try it free first.

Subscribe now

Monthly

$14USD / month

Pay today and start right away. No trial period.

  • You are charged $14 USD today
  • Then $14 USD every month until you cancel
Subscribe now

Free trial

7 days free

$14USD / month

Try everything for 7 days before your first charge.

  • A payment method is required
  • Nothing is charged for 7 days
  • Then it renews at $14 USD / month unless you cancel before the trial ends
Start your 7-day free trial

Everything included, with either option

  • Explanations · Chapter breakdowns & diagrams
  • Equations · Step-by-step derivations
  • Code · Incremental implementations
  • Flashcards & quizzes · Spaced repetition & checks
  • Chat · Q&A grounded in the paper
  • Discovery · arXiv & OpenAlex search
  • Knowledge map · Cross-paper connections
  • Learning plans · Roadmaps built from your goals

Plus Teacher across your library, audio lessons, peer reviews and agent access.

  • Your AI usage isn’t included: OpenRouter bills it separately.
  • Prices are in US dollars. Depending on your country, sales tax or VAT is calculated at checkout.
  • Each account can use the free trial once. While your tester access lasts you can only subscribe directly; when it ends, you can start the trial if you never used it.
  • Cancel anytime; you keep access until the end of the period you paid for.
  • Refunds within 14 days of a payment, with either option. See the Refund Policy.
FAQ

Questions, answered

What is OpenRouter, and why do I need a key?
OpenRouter gives you one API key for AI models from Anthropic, OpenAI, Google and many others. Structured Learning runs every AI feature on your own OpenRouter key: you choose the models and pay OpenRouter directly, with no markup from us. Creating an account and a key at openrouter.ai takes a few minutes.
How much does the AI cost per paper?
It depends on the paper’s length and the models you pick, and illustrated explanations cost more than text-only ones. You can set a spending limit in Settings, and the Usage page shows what every run cost.
Which files can I upload?
PDF, Word (DOCX) and LaTeX files up to 50 MB. Scanned PDFs work too: pages without a text layer are read with OCR.
Is there a free trial?
Yes, 7 days, and it’s optional. You enter a payment method when you start and aren’t charged during the trial. Cancel before it ends and you pay nothing; otherwise it renews at $14 USD / month. Each account can use the trial once. If you’d rather skip it, choose Subscribe now: you’re charged today and there is no trial period.
Can I cancel anytime?
Yes. Cancel whenever you like and you won’t be charged again. You keep access until the end of the period you already paid for.
What is your refund policy?
If Structured Learning isn’t right for you, ask for a refund within 14 days of a payment. The Refund Policy has the details.
Who can see my papers?
Only you. Your library is private to your account. To explain a paper, its text is sent to the AI models you choose through your OpenRouter account. The Privacy Policy explains what we store and why.
Who handles payments?
Our order process is conducted by our online reseller Paddle.com. Paddle.com is the Merchant of Record for all our orders: it handles checkout, payment security, sales tax and invoices, and answers billing questions. Your card statement will show a charge from PADDLE.NET.
Do I need to install anything?
No. Structured Learning runs in your browser on a computer, tablet or phone.

Start with the paper you’ve been putting off.

Start your 7-day free trial, connect OpenRouter and upload it. Your course starts building right away.