See exactly how Semora stacks up

Student planning apps look interchangeable on a features grid and behave nothing alike once week six arrives. Most of them are really one of three things wearing similar marketing. The first is an intake tool: something whose main job is getting coursework out of a document or a school system and into a structured list. The second is an organizing tool: something that holds deadlines, grades, and a class schedule and answers the question of what is due and what it is worth. The third is a materials tool: something that takes a PDF, a lecture recording, or a slide deck and produces flashcards, quizzes, summaries, or tutoring on that specific content.

A few products span two categories. Very few do all three well, and pricing rarely tells you which category you are actually buying. The useful move is to decide which one you need before you open a single comparison table, because a tool from the wrong category cannot be fixed with a higher tier.

Decide by naming what breaks. Write down the last three assignments you missed, submitted late, or finished at 3 a.m., and name the cause of each. If the cause was not knowing something existed, you have an intake problem. If the cause was knowing and not sequencing, you have an organizing problem. If the cause was sitting down with the material and not absorbing it, you have a studying problem. Buying a flashcard generator to fix an intake problem is the most common and most expensive mistake in this category.

• You have an intake problem if deadlines exist somewhere you do not look — a syllabus PDF, a course site, an announcement email.
• You have an organizing problem if you know what is due but keep discovering that three things are due the same week.
• You have a studying problem if the dates are handled and the material is not.
• You may genuinely have two, in which case pair one tool per problem rather than looking for one app that claims all three.

There are only three intake paths, and they are not interchangeable. A tool can read a syllabus document, connect to your school's learning platform, or make you type everything yourself. Which one a product supports shapes everything downstream, because whatever never gets in never gets tracked.

A learning-platform connection mirrors what an instructor publishes in the platform. That is genuinely useful for assignments added mid-semester, but a syllabus is the authoritative document for the things that decide your grade: the weight percentage attached to each category, the exam dates set in week one, the reading schedule, the office hours, the letter-grade scale, and the late-work policy. A large share of that never becomes an entry in a learning platform. A tool that only syncs an LMS inherits every gap in how your instructors use it. A tool that only reads syllabi will not see the assignment your professor invents in week nine. Ask which paths a product supports, and which of them are included on the tier you would actually pay for.

Semora's scanner is on the free tier and takes four paths: a camera photo of up to five pages per scan, budgeted at 10 MB combined; a PDF upload; drag-and-drop on the web app; and pasted raw text on web, up to 60,000 characters at a time. From those it extracts the course name and code, the instructor, class meeting blocks with days, start and end times, kind and room, office hours, the semester start and end dates, the letter-grade scale, and every assignment, quiz, exam, project, and reading with a due date, due time, percentage weight, and a confidence score. Importing from Canvas, Blackboard, or Moodle is a Pro feature. The current Canvas connector uses a personal access token generated in Canvas and may be disabled or prohibited by your institution. If token use is unavailable or not permitted, paste Canvas assignment text straight into the scanner.

• Does it read a syllabus at all, or does it require an LMS connection or manual entry to get anything in?
• If it syncs an LMS, which one — a Canvas-only integration is useless if your school runs Blackboard or Moodle.
• Does the connection need your institution to enable something, or can you set it up yourself?
• Which intake paths are on the free tier, and which are behind the subscription?

Every AI extraction is probabilistic. Nobody parses a scanned, two-column, badly photocopied syllabus with perfect accuracy, and any product implying otherwise is describing a demo rather than your actual course load. What differentiates these apps is not whether they make mistakes but what they do with their own uncertainty.

Silent import is the failure mode to look for. If a tool drops forty extracted items straight into your calendar with no review step, you inherit every misread date without knowing which ones to distrust, and you find out in week six when something was due in week five. Good handling looks like an editable review screen before anything is committed, a per-item signal for low-confidence extractions, a sanity check against the term dates so a date parsed as the wrong year is visible, and somewhere to park items that have no date at all instead of inventing one.

Semora splits the write in two. The course itself, its meeting times, and its grading scale are saved before you review, because those are stable and you would only have to re-enter them. Deadlines wait for your approval. The review screen shows every extracted item with per-item editing, badges anything below 0.8 confidence with "Low confidence — please verify", badges dated items that fall outside the term with "Date looks outside this term — double-check", and collects undated items in a separate "Needs a date" section that is deselected by default, so nothing dateless slips into your calendar unnoticed.

• Is there a review step at all, and can you edit each item in it before saving?
• Does it distinguish items it is confident about from ones it guessed?
• Does it check extracted dates against the semester window?
• What does it do with an item that has no date — drop it, guess, or hold it aside?

Grade tracking is where feature lists are least honest, because three very different things share the name. Some apps have none and simply track deadlines. Some ship a standalone calculator: a separate screen you feed by hand, disconnected from the assignments you are already tracking, which means double entry and a number that goes stale the moment you stop maintaining it. Some wire the gradebook to the same items the deadline list is built from, so entering a score once updates everything.

The questions that separate them are mechanical. Does it support weighted categories, or does it average everything equally as if a quiz and a final were the same thing? Can a category drop its lowest score, the way most syllabi actually work? Is there a rule for extra credit? Most importantly, does the running average divide by the weight you have attempted so far, rather than treating ungraded work as zeros — the difference between a tool that says you are at 91 percent and one that says you are failing in week three. Finally, can you edit the letter cutoffs, since a school using a 93-for-an-A scale and one using 90 will disagree about your grade.

Semora includes grade tracking on the free tier, computed from the scores and weights you enter, reflecting only work graded so far. It supports weighted categories, dropping the lowest scores within a category, and an extra-credit policy. Pro adds Grade Scale and Forecasting: custom letter-grade cutoffs matching your school's scale, and what-if calculators for what you need on remaining work to hit a target. Progress Insights, also Pro, adds trend charts, CSV export, and a print view.

• Weighted categories, or a flat average of everything?
• Does the average reflect only graded work, or does it count ungraded assignments as zeros?
• Can you change the letter-grade cutoffs to match your school?
• Is the gradebook fed by the same items as the deadline list, or a separate screen you maintain by hand?

Deadlines and study plans are different kinds of data and should not be fused. A due date is a fact published by your instructor. A study schedule is derived: a guess about when you will do the work, generated from those facts plus your availability. When a product blends them into one object, correcting a fact means rebuilding your plan, and rebuilding your plan risks losing the corrections.

The questions to ask are about rebuild behavior. Does the plan regenerate when deadlines change, or is it a one-time export you have to redo by hand? When it regenerates, does it destroy the sessions you already completed, or does the work you have done reduce what remains? Does it know your class times so it does not schedule an hour of study on top of a lecture? Does it respect a realistic daily ceiling rather than filling every waking hour? And what does it do with an item that has no due date, which cannot be scheduled backwards from anything?

In Semora, deadline and grade tracking are free and Smart Plan is Pro, so the underlying facts stay usable whether or not you subscribe. Smart Plan builds a deterministic two-week schedule from your current deadlines, sized by task type and weight. Sessions you have already completed are preserved and their minutes are subtracted from what remains, so a rebuild carries unfinished work forward without duplicating finished work. Class meetings are blocked out, and device-calendar events can be too. You control the daily minute ceiling, session length, weekday and weekend start times, and whether weekends count at all. Tasks with no due date are skipped rather than guessed at, one more reason the review screen keeps undated items visible.

• Does the study plan rebuild automatically when a deadline moves?
• Does a rebuild preserve completed sessions, or wipe your progress?
• Does it schedule around class times and a realistic daily capacity?
• Can you keep your deadline data if you stop paying for the planning layer?

This is the single best test of a course tool, because dates move constantly and every weakness shows up at once. A professor pushes a paper by a week. Four things should follow, and two things should not.

The four that should follow: the item updates in place rather than appearing twice; scheduled reminders re-fire against the new date instead of the old one; any calendar event the app created gets edited rather than duplicated; and any derived study plan rebuilds around the change. The two that should not: work you already marked complete must never come back as incomplete because a sync overwrote your local state, and items must not vanish because a platform stopped listing them — an assignment hidden by an instructor is not an assignment that no longer exists.

Semora's LMS import matches each incoming assignment on its provider-side identifier, so a re-sync updates the existing item's date, time, title, and points instead of creating a duplicate. Completion is merged rather than overwritten, so a task you checked off stays checked off. Items no longer returned by the provider are flagged as removed rather than deleted, and the server explicitly tells the client not to infer deletions from a truncated response. After a sync, reminders for every affected task are rescheduled. A task already synced to your device calendar has its existing event updated rather than a second one created.

• Move a real due date in the first week and watch what the app does with it.
• Check whether a completed item stays completed after a sync.
• Check whether the reminder now fires against the new date.
• Check for duplicates in your device calendar, which is where they usually surface.

Per-device availability is not a footnote. If you read a syllabus on a laptop, capture homework on a phone, and study on a tablet, an app that only exists in one of those places quietly becomes an app you stop opening. Check three things separately: which platforms have a real client, whether one account covers all of them, and whether editing in one place shows up in the others without a manual refresh.

Then check where the money happens, because it is not the same question. Subscriptions bought through an app store are managed in that store, and cancellation lives in your device settings rather than in the app. Some products sell on the web instead, some sell in both places at different prices, and some sell in one and honor it everywhere. Ask specifically whether a purchase made on your phone applies when you sign in from a browser.

Notification behavior also differs by platform in ways marketing pages flatten. Scheduled reminders on a phone are local notifications the operating system fires whether or not the app is open. A browser tab cannot make the same promise. If reminders are the reason you are installing something, confirm they work on the device you will actually be near.

Semora is one universal iOS app covering iPhone and iPad, plus a web app, on one account, syncing in near real time. There is no Android app and no Mac app. Pro is purchased in the iOS app through StoreKit and applies account-wide, including on the web, where the app reads and refreshes the entitlement rather than running its own checkout. Scheduled reminders are native-app notifications. Courses and tasks created or edited without a connection are queued locally, retried with backoff, and checked against the server version so a conflicting edit surfaces rather than silently overwriting.

Free means at least four different things in this category, and the differences matter more than the price. There is ad-supported free, which is complete but interrupted. There is trial free, which ends on a date and takes the product with it. There is feature-capped free, where the core works and specific capabilities are withheld. And there is volume-capped free, where everything works up to a quota.

For a trial specifically, check whether it requires a card, what it renews at, and whether the introductory rate applies to you — introductory pricing is frequently limited to first-time subscribers, so returning after a cancellation can mean paying full price immediately. For a capped free tier, the questions are what the cap counts, when it resets, and whether the capped thing is something you do constantly or a handful of times per semester. A limit on syllabus scans is a very different experience from a limit on how many assignments you can track, because you scan a few times in the first week and track deadlines every day for four months.

Ask one more question that almost nobody asks before subscribing: what happens if you stop paying in the middle of a semester. Does the data become read-only, disappear, or stay editable with only the paid automation switched off? And can you get your coursework out — a calendar export, a CSV, a print view — if you decide to move.

Semora's free tier is not a trial and does not expire: five syllabus scans per calendar month, up to four courses per semester, full deadline and task tracking, grade tracking with weighted averages, same-day reminders, and joining a Course Space a classmate invites you to. The scan quota resets at the start of each calendar month. The course cap applies when you create a new course in a semester, and re-scanning a course you already have does not consume another slot. Pro is $3.99 per month or $19.99 per year, which works out to about $1.67 per month annually — roughly 58 percent cheaper than paying monthly. Calendar sync with .ics export, CSV export, and the print view are Pro; if Pro lapses, device calendar sync pauses rather than deleting what it created.

Everything above collapses into a short test you can run in one evening, before you have committed four months of coursework to something. Do it in the first week, when switching still costs nothing.

Start with your worst syllabus, not your cleanest one. Scan the one that is a photocopy of a photocopy, or the one with the reading schedule buried in a table. Any tool handles a well-formatted PDF from a large department. What you are testing is what happens with the one that is not, and whether the app tells you when it is unsure.

Then run the moves that will happen anyway. Change a due date and see whether the app updates in place or duplicates. Mark something complete, sync, and confirm it stays complete. Enter two scores in two weighted categories and check the running average against your own arithmetic. Sign in from a second device and confirm the change you just made is there. Finally, find the export and the account-deletion path before you need them — a product that makes leaving easy is usually the one that is not counting on making it hard.

• Scan your messiest syllabus first, not your cleanest.
• Verify every date it extracted against the source document once, in week one.
• Move a date, complete a task, and re-sync, then look for duplicates and reverted checkboxes.
• Check your grade math by hand once, early, while the sample is small enough to verify.
• Confirm the free tier covers what you do daily, not just what you do in week one.
• Locate the export and the delete-account option before you commit a semester.

Semora is an intake and organizing tool with a studying layer on top, in that order. The scanner and the deadline and grade tracking it feeds are the core, and they are free. The planning and study features — Smart Plan, the Workload dashboard, Flashcards, the Focus timer, the AI Tutor, Grade Scale and Forecasting, calendar sync, custom reminder timing with quiet hours, Academic Risk alerts, Progress Insights, hosting a Course Space, and Canvas, Blackboard, or Moodle import — are Pro. That split is deliberate: the facts about your semester should not be behind a paywall, and you should be able to test whether the extraction is any good on your own syllabi before deciding anything.

It is not the right tool for everyone. If you need a Mac or Android client, if your main problem is turning lecture recordings into study material rather than turning syllabi into deadlines, or if you want time-blocked scheduling as the central product rather than a layer over a deadline list, other tools in this category are built closer to that shape.

The comparison pages here go product by product, using each company's own published material and flagging anything not directly confirmed on their own site rather than asserting it. No ratings or user counts anywhere on this site were made up, and where a company self-reports a figure it is labeled as that company's own claim rather than a verified number. That includes us: Semora launched recently and has no public rating history to cite, and we are not going to manufacture one. Pricing especially is worth checking directly at the source before subscribing to anything, ours included.