更新履歴：バグ修正と改善

Bug Fixes

• AI video lectures now work from photos and scanned documents. Previously, creating a video from an image or an image-only PDF could fail with an unhelpful message telling you to pick a different model — now we automatically read your images and build the video.
• Fixed AI Flowcharts where the steps could render as solid black boxes that hid their labels — every step now shows as a clean, readable box with its text clearly visible.

Bug Fixes

• Flashcards no longer fail on a brief AI hiccup — if the model returns an empty response, we now retry automatically (and fall back to a backup model) instead of failing your deck and using up a creation.
• Fixed a rare issue where creating flashcards could fail with an error or get stuck partway — even when your cards had already been generated. Flashcard creation now reliably saves your finished deck.
• Clearer message when an AI video lecture source has too little text — instead of wrongly suggesting your file is scanned or telling you to “run OCR,” we now simply say the source is too short and to add more detail.

Improvements

• Tidied mobile sign-up/login forms and fixed a rare Google sign-up crash.

A cleaner AI Meeting Notes page

• A slimmer, more elegant player — the bulky playback panel is now a single, compact bar, so your transcript, summary, and tasks get more room and the whole page feels calmer.
• Easier to read — refreshed the layout with more breathing room around your transcript, summary, and action items.

Improvements

• Chat now quietly rides out brief hiccups from the AI provider — if a response can’t start because the model is momentarily overloaded, we retry automatically instead of showing an error.
• Signup starts you in the free workspace — the final “How will you use Scholarly?” step now always pre-selects the free “Just me” option, so tapping Continue takes individuals straight into their personal study space. The paid Teams option stays one tap away, just never pre-picked.
• Share entire folders — folders can now be shared just like any other item. Recipients get a read-only view of everything inside, and folders shared with “Your team” show up in your team’s shared library automatically.

Bug Fixes

• Clearer messages when flashcards can’t be created — if you paste a short topic or upload a photo with no lesson content, we now explain what to do for that kind of input instead of showing PDF-only advice about “pages” and “Print to PDF.”

Introducing Scholarly for Teams

• One workspace for your whole group — put a class, department, research lab, school, or company on a single plan. Every paid feature turns on automatically for every member, and everyone works from your shared material.
• The most capable AI models — teams unlock our top tier of models across every AI feature, with admin controls to choose which model tiers the team can use.
• Weekly AI credits per member — each person gets their own weekly credits, never a shared pool, so one teammate’s heavy week never eats into anyone else’s.
• Admin controls in one place — invite by email or link, Member and Admin roles, shared source libraries, and one central bill. See Scholarly for Teams to start a team.

AI Image Playground upgrades

• High-quality mode — a new Quality selector lets paid plans generate sharper 2K images on models that support it, and shows exactly which models will use it.
• See which model is generating — while images render (and if one fails), each tile now shows the model producing it, so comparing models side by side is much easier.
• Removed the Style menu — describe the look you want directly in your prompt instead.
• Deleting an image now asks for confirmation first.
• Fixed the remaining-generations counter showing “Infinity free left” on unlimited plans.

AI now reads your PDF pages the way you see them

• Generations grounded in the actual pages — when building slides, study guides, worksheets, videos, and other content from a PDF, the AI now reads the relevant pages as full page images by default, so diagrams, equations, tables, and handwriting are read exactly as they appear instead of relying on extracted text that can drop or garble them.
• Infographics and mind maps see your pages too — they now look at your PDF’s actual pages when planning, so visual material like charts and diagrams makes it into the result.

A cleaner, more consistent look across the app

• Light mode got a big polish pass — hundreds of panels, cards, and labels across chat, recordings, podcasts, flashcards, settings, and more now have proper contrast, so nothing looks washed out or blends into the page.
• Dark mode fixes everywhere — missing dark styles on hover states, cards, and text were filled in across the app, and equations in the page editor no longer render as white strips in dark mode.

Improvements

• Print your PDFs and slides — the three-dots menu on PDF and slide viewers now has a Print option that opens your browser’s print dialog with the document ready to go.
• Pasting a link no longer uses up your free file upload — links you add as sources are treated as sources, not file uploads, so your lifetime upload credit stays available for real files.
• Added a Skip option to the welcome tour so you can jump straight into the app.

Bug Fixes

• Quiz and exam questions now test understanding, not diagram trivia — questions built from your material no longer ask you to recall an exact value or timing read off a source’s illustrative graph (like “at what time does the curve peak?”). They now focus on the mechanism and relationships the figure is showing.
• Fixed upgrading being blocked for some users — clicking an upgrade button could show “An error occurred while creating the checkout session” instead of opening the checkout page. Checkout now opens reliably, and we’ve added extra safeguards so this class of problem can’t block a purchase again.
• Fixed Home and All Content sometimes reloading more of your library than needed when switching views or filters, so large libraries stay faster and smoother.
• Fixed the welcome guide overstating how many free AI creations some accounts include — it now always matches your actual allowance.
• Fixed pacing markers like “[slow]” occasionally appearing in AI video lecture transcripts.
• Clearer help when a PDF is really just a printed document-viewer page — if you upload a “Print to PDF” of an online document viewer (SharePoint, OneDrive, or Google Docs) that was saved before the document finished loading, the file only contains the page’s web address and a page number, not the actual document. We now spot this instantly and point you to download the original file — across flashcards, quizzes, podcasts, slides, and AI videos — instead of trying for a while and then showing a confusing error.
• The Print button on the Formula Sheet Generator now works everywhere — it reliably prints just your formula sheet (not the whole page), and if your browser blocks the print window it downloads the sheet for you instead of showing an error. This fixes it on iPad and iOS, where printing used to dead-end.
• Fixed the billing and connections pages showing a bright white background for dark-mode users.
• Fixed chart labels in flashcard statistics being nearly unreadable in dark mode.
• Fixed several menus and drawers feeling sluggish when opening — they now respond instantly.
• Fixed the “last viewed” time on shared content showing an older visit instead of the most recent one.

Custom Designs are now free for everyone

• Save your own look and reuse it — on any plan — create a custom design (your style instructions plus reference images) and apply it when generating Slides, Story Books, Infographics, Mindmaps, and AI Videos. It used to be a premium-only feature; now everyone can build and use them.

Improvements

• Your most-used dialogs now open instantly — Upload Content, flashcard and quiz creation, the welcome tour, and the plans dialog no longer pause for a moment the first time you open them.
• Your folder’s name now shows on your phone — opening a folder on mobile now displays its name at the top, so you always know which folder you’re working in.
• Deleting a custom design now asks you to confirm — removing a saved design used to happen on a single tap, which was easy to trigger by accident. We now show a quick confirmation first, so you won’t lose a saved look and its reference images by mistake.

Bug Fixes

• You can now remove reference images in Custom Designs on your phone — the remove (×) button on a reference image was only visible on hover, so it couldn’t be tapped on a touchscreen. It’s now always visible on mobile, and the Save / Back buttons stack neatly instead of crowding one line.
• Saved web pages now work as study material — if you add a saved web page (for example a Google Drive or Google Docs viewer page) as a source, we now pull out its readable text automatically instead of failing. When a page is really just a sign-in or viewer shell with no content, we tell you right away to download the actual file — instead of trying and failing a minute later.

Voice Input in Chat

• Speak your message instead of typing it — there’s a new microphone button next to Send in AI Chat. Tap it and the input turns into a live recorder with a waveform that moves as you talk; tap stop and your words are transcribed straight into the message box, ready to edit or send.

Improvements

• A proper welcome on your phone — new sign-ups on mobile now get a quick, phone-friendly setup: pick what you want to make (flashcards, an AI video lecture, a quiz, or a study guide), then snap a photo of your notes, choose a file, or paste text — and Scholarly starts building it right away. Previously, phones skipped the welcome tour entirely.

Help Center

• A full refresh of every help guide — all 58 articles were checked against the current product, and 37 were updated so what you read matches what you see: the new model picker, PDF page selection in every create flow, typeset document themes, AI Meeting Notes, Deep Research file deliverables, and current video modes and lengths.
• Four new guides: Paste Text (turn pasted notes into study material), Studying from YouTube Videos, Deleting & Restoring Content, and Account, Login & Password Recovery.
• Removed outdated information across help and product pages — including references to a free trial that no longer exists and features that have been retired.

Bug Fixes

• Scanned PDFs from Google Drive now work for flashcards — a scanned or image-only PDF picked from Google Drive used to fail with a “blank or unreadable” error even when it was full of content. We now read its pages as images (the same way uploaded scans already work), so it generates cards normally.
• Clearer messages when a source can’t be read — if your Google Drive connection has expired you’re now told to reconnect it (instead of a generic “try again”), and empty files or not-yet-transcribed videos now explain exactly what to fix — across AI video lectures, slides, study guides, infographics, and podcasts. These cases also stop retrying pointlessly and refund your credit right away.
• YouTube videos no longer fail with a “no transcript” error when building an AI Video Lecture or Podcast — some videos (most often long ones) could report no transcript even though captions were available. Scholarly now recovers the transcript automatically, so generation just works without needing to open the video first.
• Fixed the billing link in payment and renewal emails — the “billing settings” link in the payment-failed and yearly-renewal emails pointed to a page that didn’t exist. It now takes you straight to your billing settings.

Improvements

• Paste Text is now a first-class way to add study material — a new Paste Text option on Home and a Paste text tab in the Add content dialog turn your pasted notes, lecture transcripts, or articles into a saved source you can study from, no file needed.
• Made it clearer that YouTube links are supported: the Upload Content tile now says so, and the Add content dialog highlights YouTube in its link section.
• Moved Study Guide to the front of the Home create row.
• Tidied up the Home page create options: the newest AI agents — AI Timeline, AI SOP, AI Flowchart, AI Lesson Plan, and AI Outline — now live under the More menu along with the AI Image Playground, keeping the main row focused while every tool stays one click away.
• Added a quick Folder shortcut and surfaced AI Chat in the Home create options, and moved Practice Exam into the More menu.
• Gave the Add content dialog more comfortable side spacing on mobile, so content no longer hugs the screen edges.
• Folders can now be renamed from the folder page — click the folder’s name (or the pencil icon next to it) in the header to edit it in place, just like renaming a PDF or flashcard deck. You can also generate a title with AI while editing.
• Redesigned the chat model picker: a focused Recommended Models list up top, with everything else tucked into a dedicated Other Models menu instead of expanding the same list — plus a full-color Gemini logo.

Bug Fixes

• Fixed AI Video Lectures coming back in English when your source material was in another language — with language set to Auto, the video now matches your source’s language (Spanish notes make a Spanish video). Picking a specific language still works as before.
• AI Video Lectures now make fuller use of their runtime on dense sources instead of over-compressing, stay closer to your material (no invented facts), and preserve your own script’s wording when you provide one in the instructions.
• Fixed the Custom Designs editor’s create button staying silently disabled — it now tells you what’s missing, and the design picker shows upfront that custom designs are a Premium feature.
• Flashcards now work with scanned or photographed PDFs from Google Drive. If a PDF has no selectable text, Scholarly reads its pages as images instead of stopping — so a scanned textbook still turns into a full deck.
• Fixed chat citations sometimes showing as raw source codes instead of clickable source badges — source links now render correctly on every chat model.
• Citations now appear as clickable source badges while the answer is still streaming — previously they showed as raw citation text until the response finished.
• Fixed AI Video Lecture captions appearing as huge paragraph blocks instead of readable subtitle lines.
• Fixed the Build Flashcards button appearing off-center in the practice exam “What would you like to study?” empty state.
• Fixed the model picker in chat not showing a hover highlight in dark mode.
• Fixed Claude Sonnet 5 erroring out when Thinking was turned on in chat — it now thinks and answers normally.
• A model’s thinking summary now appears above its answer instead of below it.

AI Images

• Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) is now the default image model in the AI Image Playground, chat image generation, Story Books, AI Video visual generation, and Scholarly’s shared image tools.
• GPT Image, Grok Imagine, and Nano Banana 2 remain available in the Playground when you want to compare styles or generate alternate versions from the same prompt.

AI Agents

• Added AI Timeline, AI SOP, AI Flowchart, AI Lesson Plan, and AI Outline — five source-grounded PDF agents available from Home, More, chat, upload, and folder create flows.
• The New menu and the home page create options now show a one-line description of what each tool does — so it’s clear what each one builds from your material.
• Added a new AI Agents page and help guide explaining how Scholarly’s agents read your sources, use multiple models, research the web, and run code to build your study materials.
• Added step-by-step help guides for AI Timelines, SOPs, Flowcharts, Lesson Plans, and Outlines.

Polished PDF Documents

• Study guides, worksheets, lesson plans, outlines, SOPs, and timelines now generate as professionally typeset PDFs — the kind of clean, print-ready handout a professor would give out, with proper headings, callout boxes, tables, and math.
• Choose a design theme — Professional (default), Academic, Modern, or Playful — right in the create screen to match the look you want.

Bug Fixes

• Fixed a chat draft you couldn’t send (for example, a message blocked after you hit the daily chat limit) carrying over into the Home page chat box.
• Fixed the Home page Content count sometimes not matching the number of items actually listed below it.
• Fixed AI Video Lectures that could fail after several attempts on longer PDFs — if the fast draft model gets stuck, we now automatically retry on a more capable model so more lectures finish successfully.
• Fixed an issue where chatting with Claude Haiku 4.5 returned an error instead of a reply.
• Fixed an error that could interrupt an AI chat after it switched between AI models mid-conversation.
• Improved read-aloud reliability and overall app stability during traffic spikes.
• Fixed the Explore page failing to load some shared materials.
• Fixed chemistry equations in chat sometimes showing raw underscores (like “O_2” or “C_6H_12O_6”) instead of proper subscripts (O₂, C₆H₁₂O₆).
• Fixed podcast transcripts and subtitles occasionally showing raw citation data instead of clean text (the audio was always fine).
• Fixed a PDF study guide issue where collapsible sections could be hidden, leaving a partial or blank guide even though the full text had been generated — all sections now show.
• Fixed Research occasionally finishing with an internal status message in place of your report; these now retry and refund the credit instead of saving a broken report.
• Fixed recording summaries inventing a full lecture from non-lecture audio (a greeting, a test clip, or background noise) — these now show an honest note instead.
• Fixed quizzes & exams where the True/False answers could come out almost all “false”; the answer key is now balanced so the test actually checks your understanding.

Improvements

• Sharper AI generation across the board: flashcards and quizzes phrase questions to test understanding (not source page numbers), AI Video Lectures stay in your chosen language, slide decks honor a requested slide count, and worksheets & study guides stay grounded in your own material.
• Voice chat now points you to the right Scholarly feature (like AI Video Lectures) instead of suggesting outside apps, and handles scanned PDF pages more gracefully.

Improvements

• Redesigned the AI model picker in the create tools (flashcards, slides, study guides, AI video lectures, infographics, mind maps, and more): models are now grouped into Recommended models up top with everything else tucked under Other models, so it’s easier to choose. Picking a specific model is a Plus feature.
• Widened the More create menu so tool names like “AI Story Book” and “AI Meeting Notes” show in full instead of getting cut off.

Mobile

• A big mobile overhaul across the whole app — phone-friendly layouts from the landing page and signup through your home screen, study tools, players, and pricing.
• Create flows now start with your phone’s own sources: snap a photo or scan your notes, pick a screenshot from your photo library, or grab a file — instead of a desktop drag-and-drop box.
• Bigger, easier-to-tap buttons everywhere, and menus, dialogs, and the plan picker now use the full screen on phones instead of cramped pop-ups.
• Fixed tapping a field zooming the page in, the mobile menu hiding half its links, and the home “Create with AI” buttons being too small to press.
• PDFs now fit your screen on phones instead of running off the right edge, so you can read the whole page without scrolling sideways.
• Cleaner content-page headers on phones — just the Scholarly logo (no cramped title bar), with content view switchers (PDF, video, recording) as tidy icon tabs.
• Balanced, centered podcast player controls on phones (they were sitting off-center before).
• Fixed timestamps like “Research · 10h ago” wrapping onto a second line in your content lists.

AI Slides

• Added Kimi K2.7 Code for AI Slides and made Gemini 3.5 Flash available in select AI creation flows, giving premium members more model options for structure, reasoning, and source handling.
• Improved AI Slides reliability for long generations so models have more room to finish and are nudged to finalize the deck instead of spending the last steps on visual polish.

Help Center

• Five new help guides for features we’d shipped without docs: Custom Designs, AI Story Books, AI Spreadsheets, AI Study Pack, and AI Essay Writer.
• Refreshed the Deep Research guide for the new file-creation flow (study guides, comparison tables, CSVs, timelines) plus inline source pills, embedded images, and direct PDF/Word/PowerPoint attachments.
• Refreshed the Choosing an AI Model guide to cover the new Recommended badges, in-picker model descriptions, and the fully browsable picker on free plans.
• Updated the Flashcards guide for the PDF page selector that now appears in every AI create flow, and for 1:1 question-bank import from CSV, Quizlet, Anki, and Excel files.
• Updated AI Video Lectures and Settings for the unified Custom Design picker and Settings → Scholarly AI → Custom Designs management.
• Refreshed Home & Library with the new Blank vs Create with AI tabs in the New menu.
• Removed outdated guides and dead links from the help index and sitemap.

Bug Fixes

• Fixed AI Story Books sometimes finishing with nothing to show on long, multi-topic requests — the builder no longer gets stuck after using up its illustrations and now completes the book.
• AI Video Lectures now stop with a clear “we couldn’t read this file” message and refund the credit when a source has no readable material, instead of creating an off-topic lecture.
• Chemical formulas in AI Chat (like C₆H₁₂O₆ or 6CO₂) now show proper subscripts instead of stray underscores, both while the answer is typing out and after it finishes.

Improvements

• AI Meeting Notes now warns you the moment it stops hearing your microphone, so you don’t record a whole session of silence before finding out.
• Clearer message when an uploaded file has no readable lesson text (for example a blank web-page export), with steps to re-upload the real content.
• AI Video Lectures now start with background music off, and the setup screen includes a visible Quick / Standard / In-depth length control.
• Deleting an account no longer requires a long written explanation, and there’s now a clearer “Not what I expected” reason option.
• Made large flashcard decks, long chats, the sidebar, and the full content library feel smoother by reducing hidden work, unnecessary re-rendering, and heavy animations.

Bug Fixes

• Fixed a crash that could blank out a page — most often study guides on the PDF page, but also videos, the home screen, and other pages — when your browser was automatically translating Scholarly (most common on Chrome for Android). Pages now stay stable while translation keeps working.
• Fixed flashcards sometimes refusing to build a vocabulary or translation deck from a source written in a single language — for example learning a Japanese book’s words with English meanings on the back. These decks now generate as expected.
• Fixed AI Video Lectures failing when you asked for a much longer video than the source could fill (for example a 30-minute lecture from a few pages). It now produces the best complete lecture the source supports instead of timing out.
• Fixed the PDF page selector not appearing in the flashcards setup when you picked a PDF from your library — it now shows for uploaded and saved PDFs alike, and sits above the advanced options.

Improvements

• Added a Recommended badge to the AI model picker for flashcards, AI video lectures, slides, and more, so it’s clear which model gives the best results — with a short description under each model (like “fast and lightweight” or “clear, well-paced lectures”).
• Premium members now automatically get the recommended model when creating with AI, no extra setup needed.
• You can now open the model picker to browse every available model even on the free plan; choosing a different model is still part of premium.
• Removed the study streak counter from the home page and study screens. Scholarly is about making your own material more useful — not daily-streak pressure — so we’ve dropped it to keep the focus on understanding.
• You can now choose which PDF pages to include in every AI create flow — flashcards, quizzes, slides, AI video lectures, podcasts, study guides, worksheets, infographics, mind maps, story books, and Study Packs — so you can focus a generation on just the chapter or section you need.

Improvements

• Made Home and the sidebar load your library faster by fetching recent and pinned items first, while All Content loads the rest as you scroll.

New — Deep Research can build files for you

• Deep Research can now create downloadable files — a study guide, a comparison table, a dataset (CSV), a timeline, or a one-page summary. Pick “Add a file” under the question box before you run your research, no special wording needed.
• Any files Research makes are now easy to spot alongside your report — ready to open, download, or save to your library.

New — Richer research reports

• Research reports now include relevant images — diagrams, figures, maps, and photos pulled in alongside the writing to make reports clearer and easier to follow.
• Inline source pills — the site behind each claim now shows as a small clickable chip right in the text, so you can see and check sources as you read.
• Cleaner, more readable formatting — more spacing between paragraphs, clearer section headings, and better-styled quotes and images.

New — Start from blank

• You can now create empty files to fill in yourself — a blank flashcard deck, a blank spreadsheet / table, and a blank PDF — right from the New menu.
• The New menu is now split into two tabs, Blank and Create with AI, so it’s easier to find what you want to make.

Bug Fixes

• Fixed AI Video Lectures failing to play on iPad and iPhone with a “video source failed to load” error — lectures now load and play reliably on Apple devices.
• Fixed some valid PDFs being rejected as “not a valid PDF” before you could create flashcards or upload them — these PDFs now upload and process normally.
• Fixed AI Story Books occasionally failing partway through on longer stories — the illustrated book now finishes reliably.
• Fixed uploading a question bank (a CSV, Quizlet, Anki, or Excel export with Question/Answer columns) only turning into a fraction of its rows — every question is now imported as its own flashcard, word-for-word, even for banks with hundreds of questions.

Improvements

• Renamed the Page create option to AI Essay Writer across Home, the New menu, and folders, so it’s clearer what you can make with it.

New — Custom Designs

• Save your own look as a Custom Design — add your style instructions and reference images once, then apply it whenever you generate Slides, Story Books, Infographics, Mindmaps, and AI Videos, so everything comes out in your style (available on paid plans).
• Create and manage your saved designs anytime from Settings, under Scholarly AI.

Bug Fixes

• Fixed AI Chat sometimes writing a chemistry formula with stray underscores (like the glucose formula C_6H_12O_6) instead of clean subscripts, and occasionally slipping a word from another language into an otherwise-English answer.
• Fixed mind maps and infographics showing up as a generic “PDF” in your library — they now carry their own “AI Mind Map” and “AI Infographic” labels and icons.
• Fixed the daily quiz limit’s “resets in” countdown showing the wrong number of hours.
• Fixed a rare crash on the video page that could happen when your browser automatically translated the page (most common on mobile) — the lecture transcript, summary, and checkpoint questions now stay stable instead of blanking out.
• Fixed uploading a file to create flashcards sometimes failing with a generic “Failed to upload” message — you now see the real reason, and if you’ve hit your free upload limit you get a clear option to upgrade instead of a dead end.
• Fixed a PDF occasionally disappearing from the upload list with no explanation when it couldn’t be read (for example after being moved, renamed, or re-synced by iCloud, Google Drive, or OneDrive) — you now see an inline message telling you to reselect it.
• Fixed PowerPoint and document uploads from iCloud, Google Drive, or OneDrive sometimes failing before the file reached Scholarly — uploads now stabilize the selected file first, and if your device needs you to reselect it, the message says that clearly.
• Fixed valid PDFs sometimes being reported as “corrupted” while the PDF reader was still loading — the app now asks you to refresh and try again instead.

Improvements

• The Study Pack “Generate” button now opens your upgrade options when you’re out of free AI creations, instead of looking clickable but doing nothing.
• Added clear Light / System / Dark labels next to the icons in the theme switcher (Settings → Theme).
• On phones, wide tables in document summaries, study notes, research reports, and help articles now scroll sideways instead of getting cut off at the right edge.
• Made key controls easier to tap on phones — including the Generate button in the image Playground and the dismiss button on the community card — and the delete-image confirmation now fits the screen properly.

Bug Fixes

• Fixed the fullscreen button on AI video lectures triggering an error on some older mobile browsers — tapping fullscreen now works (falling back to your phone’s native full-screen player) instead of failing.
• Fixed AI Podcast, Slides, Study Guide, Worksheet, and Story Book creation sometimes failing with an “Invalid data” error when you wrote longer custom instructions alongside a chosen style — the instructions box now reflects the real space available, so your creation always goes through.
• Fixed AI Chat occasionally saying “the PDF is ready” when no file actually appeared — a chat-built PDF is now always saved to your library and openable, even when its text can’t be fully indexed right away.
• Fixed quiz and practice-exam questions sometimes displaying broken text like “ext{...}” instead of chemistry and math notation (for example FADH₂) — those formulas now render correctly in question stems and answer choices.
• Fixed AI Video Lectures sometimes presenting sections out of order — a lecture now opens with its introduction and walks through multi-step processes in sequence, and the chapters and transcript match what’s actually narrated.
• Fixed AI Podcasts occasionally reading source-reference codes out loud mid-sentence — the narration now stays clean and never voices citation tags or stray markup.
• Fixed AI Podcasts and document summaries sometimes inventing content from a blank or scan-only PDF (or a file that didn’t load) — when there’s no readable text, you now get an honest “we couldn’t read this file” note asking you to re-upload, instead of a confident summary or episode about nothing.
• Fixed AI Video Lectures sometimes coming back in the wrong language when your source was written in another language — a lecture now always narrates in the language you picked.
• Fixed Deep Research reports occasionally showing garbled numbers and date ranges, or placeholder “example.com” links — figures now display correctly and every cited source points to a real page.

Improvements

• Flashcards now focus on the concepts you need to learn — when you upload a research paper they skip trivia about the journal, license, or study method, and they no longer ask about where a picture sits on the page.
• The voice tutor now answers from the PDF, video, or recording you have open instead of giving generic advice — ask about “this document” or a specific page and it reads your actual source first.
• Story Books now teach the real idea from your material, pitched at your level — so an exam-prep topic comes back as a genuine explainer, not just a cute story.
• Redesigned the folder view to match the rest of your workspace: your folder’s content is now front and center as a clean list, the “Create with AI” tools sit right at the top so you can make a video, deck, podcast, and more straight into the folder, and the folder chat is now a toggle you open when you want it instead of taking up the whole middle of the screen.
• AI Video Lectures now have a paid-plan Scholarly logo toggle, so subscribers can generate videos without the logo or wordmark in the frame.
• Practice quiz and exam questions now focus on understanding instead of trivia — they no longer ask which tissue, animal, lab, or year a finding was first discovered in, and instead test the concept, mechanism, or relationship behind it.
• The upgrade screen is now clearer about what you get and what you pay — it shows the full price up front (total, per-month, and how much you save), so there are no surprises when you reach checkout.

Bug Fixes

• Fixed a sign-in error that could appear the first time you signed in with Google.
• Chat replies and flashcard generation no longer fail on a brief network hiccup — they now retry automatically.
• Word and PowerPoint files that are too complex to convert now fail quickly with a clear message instead of hanging for several minutes.
• Fixed large and poster-format PDFs (scanned textbooks, engineering drawings, oversized scans) failing to process during upload — their pages now render reliably, and one heavy PDF can no longer slow down or interrupt other uploads.
• Fixed flashcards being wrongly rejected as a “title page or table of contents” when you uploaded a study guide, review sheet, or list of key terms — those term and topic lists are exactly what flashcards are for, and now generate a real deck instead of an error.
• Fixed Deep Research ending with a generic “couldn’t finish” error when a file you referenced wasn’t attached or searches came up empty — it now stops hunting and gives you a clear, honest result telling you exactly what was missing and what to do next, instead of running for minutes and failing empty-handed.
• The login screen now shows a clear message when it can’t reach the server, instead of a confusing technical error.
• Fixed a duplicate arrow appearing on some call-to-action buttons across the feature and comparison pages.
• Fixed AI Chat sometimes getting stuck or failing when asked to build a file — PDF reports, Word documents, PowerPoints, and charts now generate reliably on the first try.
• You can now open a PDF or document that Chat saved to your library in its own tab again, straight from the chat — the “open in new tab” button was missing for AI-generated files.

Improvements

• Deep Research now lets you attach PDFs, PowerPoint, Word, and text files — not just images. Your real source documents are read directly during the research instead of being left out.
• Refreshed the welcome screen to spotlight the study tools students use most — flashcards, AI video lectures, mock exams, the AI tutor, and podcasts — with everything else rolled into a single “and so much more.”
• AI Chat now shows every step it takes — including building and reading files — so you can follow exactly what it did.

Sharing

• Sharing your work is now front and center — saving a video, PDF, flashcard deck, podcast, recording, or research report opens a quick share panel so you can send a link to a classmate in one tap.
• Opening a shared link feels smoother: shared pages now show a calm, branded loading screen instead of a blank page with a spinner, and the “save to your account” prompt now waits until you’ve started reading instead of popping up right away.
• Cleaned up shared content pages: they now look like the focused study view instead of showing a marketing footer, with a tidy “Sign up” button in the top right.
• On a shared PDF or video, the summary now loads and displays on its own instead of showing a “generate” form to visitors who aren’t signed in.
• The “Create” tools (podcast, video lecture, slides, flashcards, quiz, and more) on a shared link now invite you to sign up first, so a tap always leads somewhere instead of quietly doing nothing.

New — AI Story Book generator

• Turn an idea — or your own PDFs and notes — into an illustrated children's story book. Scholarly writes a warm, age-appropriate story and illustrates every page, keeping the same characters and look throughout, then delivers it as a page-by-page book in your library.
• Choose an art style: Classic Storybook, Soft Watercolor, Bright & Playful, or Dreamy Pastel — and add your own instructions to steer the story.

Help Center

• New guide: AI Image Playground — the full walkthrough for the new image workspace, including how to pick a style, an aspect ratio, generate with multiple models at once, attach reference images, and how the Playground differs from generating images inside chat.
• Refreshed the Video Modes guide for the new Editorial, Retro, and Freestyle modes — the picker is now six modes, and each one has its own breakdown of when to reach for it.
• Updated the AI Video Lectures guide and the matching landing pages with the same six-mode picker.
• Refreshed the Choosing an AI Model guide with DeepSeek V4 Flash, the new fast model that’s now the default for AI Video Lectures.
• Updated the AI Study Guides guide for the new themeless, print-ready PDF format — the theme picker is gone, and every study guide now renders as a clean, professional document built to study from and print.
• Cross-linked the AI Image Generation guide with the new Playground so it’s clear when to reach for which.

Bug Fixes

• Fixed AI quizzes and practice exams occasionally generating their questions in a different language than your study material — for example, English flashcards producing questions in another language. Questions now reliably match the language of your flashcards.
• Fixed a bug where creating flashcards could fail with “the content could not be processed” on perfectly valid material — if one AI model declines, it now automatically retries with a backup model instead of giving up.
• Fixed a rare crash that could stop a PDF study guide from displaying.
• Fixed flashcards from scanned or image-based PDFs (like photographed or scanned textbook pages) being wrongly rejected as having “no study content” — those pages are now read with image understanding, so you get a real deck instead of an error.
• Fixed flashcard generation failing with a “content too large” error when you asked for a very large number of cards (for example “400+”) from a short source — it now returns a complete, focused deck instead of throwing away the whole generation.
• Fixed Deep Research occasionally failing without producing a report — it now reliably finishes with a report built from what it gathered instead of ending empty-handed.
• Fixed AI Video Lectures sometimes showing the wrong length (like “0:05”) or appearing not to play while a large video was still loading. The player now shows the correct length right away and, if a video is slow to load, gives you a clear “Retry” instead of an endless spinner.
• Fixed AI Video Lectures ignoring your requested narration voice — a video set to a female voice no longer narrates in a male one, and videos in other languages now use a matching native voice when one is available.
• Deep Research now sticks to the format you asked for — if you request a slide-by-slide outline, a table, or specific sections, it delivers that instead of defaulting to a plain prose write-up.

Community

• You can now join our Discord community straight from Scholarly — find it in the sidebar, under Settings → Help & Support, and right after you finish a study session, quiz, or exam. 150,000+ students are there to swap tips and help each other get unstuck.

Improvements

• Large PDFs and slide decks now open and scroll more smoothly by only rendering nearby pages and thumbnails instead of loading every page at once.
• Opening study pages and focus-mode sessions feels faster, especially on large pages or quizzes, thanks to less hidden chat and question content being mounted in the background.
• Improved app responsiveness while background AI creations are being checked and while settings are being saved.

Improvements

• Study guides now generate as clean, professional, print-ready documents — plain, easy-to-read pages built for studying and printing, instead of a styled theme.
• You can now add reference images in the AI Image Playground — upload (or paste) up to 4 of your own images and describe how to edit or remix them. Works with GPT Image, Grok Imagine, and Nano Banana Flash.
• The AI Image Playground can now generate with multiple models at once — pick any combination of GPT Image, Grok Imagine, and Nano Banana Flash, and it creates one image per model from a single prompt so you can compare styles side by side.
• AI Video Lectures now generate with a faster new default model (DeepSeek V4 Flash), available alongside the existing options in the video customization step.
• AI Video Lectures (Editorial, Retro and Freestyle) now follow a clearer order — they introduce the topic first, keep multi-step processes in sequence, and cover every part you asked for instead of opening with a summary or saving the basics for last.
• In the AI Image Playground, opening one settings menu (style, model, or aspect ratio) now closes the others so they no longer overlap.
• The cookie banner now lets you Reject non-essential cookies, not just accept them.
• The navigation menus on our website now open and close with a smooth, snappy animation instead of popping in abruptly.
• Improved app responsiveness during periods of heavy server load so everyday actions like opening your study content stay fast.

Bug Fixes

• Generated slides, study guides, infographics, and AI video lectures now display math and equations correctly — formulas like fractions, sums, and Greek symbols render as real notation instead of showing raw $...$ code.
• Fixed flashcard decks occasionally failing to open during brief spikes in server load — they now retry smoothly instead of erroring.
• Fixed AI Video Lecture sometimes ending with a confusing technical message after a momentary processing hiccup. It now shows a clear "temporary issue, please try again" message and your credit is refunded.
• Fixed asking the AI chat to make flashcards from a source it can't find — it now asks you to pick the right source instead of failing the message.
• Fixed the email that lets you know a research report or AI video lecture failed to generate — including the note that your AI creation was refunded — occasionally not being sent.
• AI quizzes and practice exams now show math symbols correctly in the answer review — the “Your answer” and “Correct answer” lines no longer display raw formatting codes.
• Picking an AI Podcast style now applies on the first click instead of needing a second tap.
• Flashcard decks set to “Low” now reliably produce a focused, key-concepts deck instead of occasionally coming back too small on long sources.

New — Editorial & Retro AI Video Lecture styles

• Added an Editorial style — a sleek, designed motion-graphics look with kinetic typography, animated charts and diagrams, and clean transitions.
• Added a Retro style — an 80s synthwave / arcade look with neon grids, glowing type, and arcade-style counters. Same material, re-taught with a totally different energy.
• Pick either in the video customization step alongside Standard, Math, Kids, and Freestyle.

New — AI Image Playground (Beta)

• Introduced the AI Image Playground — a dedicated space to turn a prompt into study visuals, diagrams, and illustrations. Pick from three image models (GPT Image, Grok Imagine, and Google’s Nano Banana Flash), choose an aspect ratio, and generate several at once.
• Choose an image style — Educational, General, Photorealistic, Diagram, Infographic, or Artistic — so the same prompt can render as a clean study visual or whatever else you need. Your style, model, and aspect ratio are remembered for next time.
• Your images keep generating even if you refresh or leave the page, and you can open any result full-screen, download it, or remove it. Open it from your home screen.
• If a prompt can’t be turned into an image because it goes against our content guidelines, the Playground now explains why up front so you can quickly rephrase — instead of a generic failure.
• It’s in beta, so daily usage is limited on every plan while we tune quality and speed.

Improvements

• The AI Image Playground now stands out with a New badge in the Home create menu, so it’s easy to spot and try.
• The chat activity trace now describes what actually happened — for example “Read a webpage,” “Read a YouTube transcript,” or the name of the file it read or saved — instead of generic labels like “Read from your library” or “Read file.”
• Freestyle AI Video Lectures now follow a clearer order — they introduce the topic up front, build through it in sequence, and keep multi-step processes together instead of splitting them across the start and end of the video.

Bug Fixes

• Fixed creating a new Page from Home sometimes leaving you on Home instead of opening the page — the new page now reliably opens right away. The same fix applies when duplicating a page or PDF.
• Improved reliability of AI generation (including flashcards) when the AI model briefly times out — these momentary timeouts are now retried automatically instead of failing the request.
• Flashcard decks now respect the number of cards you pick — choosing a smaller deck (like “Low — key concepts only”) no longer produces hundreds of cards on long sources.
• Chat no longer shows a stray “Feedback noted” card when you didn’t actually give any feedback.
• On Deep Research, the limit message no longer says “daily” on plans where your free AI creations don’t reset each day.
• Scanned or image-only PDFs from Google Drive now work for AI Video Lectures. A photographed or scanned book PDF (no selectable text) used to fail with “no extractable text” — we now read the pages as images, so it generates a video instead of dead-ending.
• Deep Research no longer runs for a long time and then comes back empty. On some topics the research could keep searching without ever writing up its findings and eventually fail with nothing to show. It now always turns the sources it gathered into a finished report, and your research is titled the moment it starts.

Bug Fixes

• Fixed large flashcard decks sometimes ending up with no practice questions in study mode — a single oversized batch no longer wipes out the questions generated for the rest of the deck, and very dense decks are now split and retried so questions come through.
• Fixed a false “Failed to save message. Please try again.” error that could appear in chat even though your message had actually been saved.
• Fixed your settings (model picks, sidebar layout, theme) occasionally not saving when you changed several in quick succession.
• Fixed a rare error when opening a PDF whose text had just finished processing.
• Picking a Google Drive file that’s too large for your plan is now caught right away — with the file’s size, your limit, and an upgrade option — before the flashcards, video, slides, or podcast even start, instead of failing partway through.
• Clearer message when a Google Drive PDF can’t be read because it’s a scanned or image-only document, and AI Video Lectures no longer keep retrying a source they can’t read.
• Diagrams and files created in chat now open in a preview panel next to the conversation instead of a new browser tab.
• Fixed AI quizzes and practice exams occasionally including an off-topic question unrelated to your material.

Improvements

• AI quizzes and practice exams now test your understanding — they focus on concepts and how things connect, instead of trivia like dates, names, or who-did-it-first.
• Freestyle AI Video Lectures now follow a clearer start-to-finish structure, introducing the topic up front and avoiding repeated sections.

Mobile

• On phones, AI Video Lecture practice questions now appear in full below the video instead of being squeezed into — and clipped by — the small video frame, so the question and every answer stay readable and easy to tap.
• Wide tables in chat answers now scroll sideways on phones instead of having their right-hand columns cut off the screen.
• More phone polish across the app: the “Create” shortcuts in chat stack to full width, long file names attached in chat no longer run off the message, flashcard and focus-mode hints now sit above your answer choices instead of covering them, and buttons in Deep Research, flashcard and quiz creation, the AI Image window, exam dialogs, and the upload window are bigger and easier to tap.

New

• Added a Freestyle video mode to AI Video Lectures — alongside Standard, Math, and Kids, this new mode lets the AI choose the best visual style and animation approach for your specific material, instead of following a fixed look.
• Added Create AI Image — a new shortcut on Home and in the “Create” menu. Describe the image you want, optionally add reference images from your device or your library, and it’s generated for you right in chat.

Improvements

• AI Video Lectures now have a Background music toggle — keep the soft music bed under the narration, or turn it off for narration over silence.
• Home now keeps your recently used AI tools at the top — the last few you opened (like AI Podcast, Deep Research, or Slides) appear in their own row so you can jump back into your favorites faster, with everything else just below.
• When a pasted link can’t be added — a private Google Doc, a page behind a school login, or a YouTube video without captions — the upload window now shows a clearer “here’s how to fix it” card with one-tap options to pick the file from Drive, upload it directly, or paste a public link.
• Polished the contrast of subtle controls in light mode — the clear-search button in the search bar and the “current folder” checkmark when moving items are now easier to see.
• Improved the readability of subtle controls and icons across flashcard study and the page editor in light mode.
• Cleaned up the sign in, sign up, and password pages — clearer headings and a crisper “Continue with Google” button for a sharper, more legible look.

Bug Fixes

• Fixed recording warnings — the “another recording is active” notice and the 3-hour limit warning — being hard to read in dark mode.
• Fixed the upgrade prompt showing a confusing “You’ve reached an AI Chat limit” message when you simply picked a premium AI model (for flashcards, slides, chat, and more). It now clearly explains that premium models need an upgrade, even if you haven’t hit any limit.
• Fixed the “Explain why” help in quizzes showing a vague “explanation limit” — since it uses the AI assistant, it now correctly tells you when you’ve reached your daily AI Chat limit.
• Fixed a generating item (such as a Deep Research report) briefly appearing twice in your Home list while it finished.
• Fixed Deep Research briefly showing “Opening research…” and a “limit reached” notice at the same time right after you started your last report of the day.
• Flashcard generation no longer fails with a confusing “try different files” message during brief AI service hiccups — it now retries automatically and only points to your files when they’re genuinely the issue.
• Fixed videos — including AI video overviews and uploads like .mkv, .avi, and .wmv — being wrongly reported as an “unsupported format” on the Transcript tab.
• Fixed a blank screen that could appear when opening a PDF study guide.
• Fixed the PDF viewer failing to open on some older browsers.
• Fixed flashcard highlights occasionally failing with a “servers are having trouble” error during a brief hiccup — it now recovers cleanly.
• Creating flashcards from chat no longer fails when the assistant refers to a source by the wrong kind of link — it now uses your other valid sources instead.
• Google Drive PDFs that fit your plan’s file-size limit are no longer wrongly rejected as “too large” when creating flashcards, slides, videos, or podcasts — reading a file from Drive now allows the same size you can upload directly.
• Google Drive files that are too large or can’t be read now show a clear message naming the file and the reason, instead of a generic error.
• Creating or submitting an exam on a shaky connection now tells you it’s a connection problem (and keeps your answers) instead of showing a generic error.
• Fixed flashcard decks sometimes failing to open with a “private or doesn’t exist” message even when you were signed in — the page now retries and shows the correct message.
• Opening the Exam tab on a brand-new deck no longer bounces you to the Questions tab — it now shows your questions being prepared and opens your exam options automatically once they’re ready.

Help Center

• Refreshed the help center top to bottom: Getting Started, Choosing an AI Model, AI Chat, and Home & Library all now cover the recent product changes — the Home chat box model picker, provider logos in the model menu, inline file preview, the All Content tabs, and the new folder tile grid.
• Updated the landing page and FAQ to highlight worksheets, mind maps, study guides, infographics, editable library files, video modes (Standard / Math / Kids), and the new ability to pick which AI model generates your content.
• Cleaned up dead and inconsistent help URLs across the blog and sitemap so links no longer 404, and help articles use one canonical address each.

Improvements

• The “Create” shortcuts on your home screen have a fresh look, and every content type — podcasts, slides, study guides, worksheets, mind maps and more — now has its own distinct color so they’re easier to tell apart at a glance.
• Studying flashcards feels much snappier — flipping a card and moving to the next or previous one is now instant, with no built-in delay between actions, and the progress bar above the card is now green.
• The attached-source chips in the chat box now fully respect your system's “Reduce Motion” setting — they fade in and out instead of springing when it's on.
• Notes and other text files now make better use of a wide screen when reading or editing — the document widens automatically as you open or close the chat and sidebar, instead of being squeezed into a narrow column with empty space on either side.
• Files linked in a chat (notes, code, CSVs, text, images, and more) now open in the preview panel right next to the conversation — the same way PDFs do — instead of opening in a new tab, so you can read them without leaving the chat.
• The upgrade and warning banners while studying, on files, and in research now look cleaner in dark mode — a softer, more legible amber with a clearer icon instead of a muddy fill.

Bug Fixes

• Fixed the quiz upgrade notice sitting off to one side with empty space next to it on the Questions tab once you've used your daily quiz questions — it's now centered.
• Studying no longer interrupts you partway through — flashcards and quizzes won't take over the screen with an upgrade prompt mid-session anymore; any prompts now wait for a natural stopping point.
• Fixed creating flashcards from an image opened in the document viewer — it no longer fails with a confusing “not a PDF” message, and your image is used as the source instead.
• Fixed flashcard generation occasionally failing after a long wait when the chosen AI model couldn't handle your sources — it now automatically falls back to another model so your cards still get created.

Improvements

• Billing and cancellation screens now use your account's saved plan details directly, so they open with fewer extra subscription checks.
• Home now loads the Jump back in flashcard review shortcuts faster by only loading the few decks shown there.
• Redesigned the All Content page with a single tabbed header (All, Pinned, Shared, Deleted) that now filters your folders too — so the Pinned, Shared, and Deleted views no longer show folders that don't belong, and deleted folders now appear under Deleted.
• Moved the content type filter next to the tabs and added a badge showing how many filters are active.
• The Slides view now has zoom controls — zoom in and out (or pinch / Ctrl+scroll) to read fine detail, and click the percentage to snap back to fit. You can also collapse the slide thumbnail rail for a wider slide.
• Simplified AI Slides and AI Video Lecture setup by keeping model choices inside the regular AI Model selector.
• Redesigned folders on the All Content page to match Home — a clean grid of folder tiles with an AI Organizer shortcut and a quick “New folder” tile that's always available, replacing the old list view and empty “No folders yet” placeholder.
• You can now pick which AI model to use right from the Home chat box, and the model picker now shows each provider's logo so models are easier to tell apart at a glance.
• Modals, tabs, and menus across the app now open and close more smoothly — their animations are lighter and GPU-accelerated, so they feel snappier with less stutter on lower-end devices, and they now fully respect your system's “Reduce Motion” setting.

Bug Fixes

• Create with AI (flashcards, quizzes, slides, podcasts, video lectures, and more) now opens to your Library so you can pick existing material right away, instead of always starting on the Upload tab. It only opens on Upload when your library is empty.
• Fixed the AI model picker in chat feeling laggy — it now opens and, especially, closes instantly and smoothly.
• Fixed the flashcard autoplay scrubber knob always showing — it now appears only when you hover the progress bar.
• Fixed AI video lectures that wouldn't play — longer videos in particular now stream and start faster, with an automatic retry if a video has trouble loading.
• Podcasts, recordings, PDFs, images, and file downloads now load directly and noticeably faster, and large files (longer podcasts and bigger PDFs especially) that previously failed to open now work, with an automatic retry if anything has trouble loading.
• Fixed creating a folder occasionally failing silently (the folder wouldn't appear) when a sync retried or you had Scholarly open in two tabs.
• Fixed flashcard generation sometimes failing right away on large or dense documents instead of automatically retrying with smaller sections.
• Fixed file uploads occasionally erroring with a “still processing” or “processor busy” message — these temporary conditions now retry automatically instead of showing an error.
• Fixed adding a private, deleted, or unavailable YouTube video showing a generic failure — you now get a clear message explaining the video can't be accessed.
• Fixed the “Generate audio again” button on a research report, which previously always failed — it now regenerates just the narration for that report.
• Retrying a research report that grew too large now explains why and suggests starting a more focused session, instead of silently re-running and failing again.
• Fixed podcasts that could finish as a very short, near-empty clip while still showing as “completed” — the app now detects this and regenerates the episode instead of leaving you with an unusable file.

Help Center

• Added two new help guides: Library Files (saving and editing the text, notes, code, CSVs, and diagrams you create with AI) and Worksheets (printable practice sets, quizzes with answer keys, and blank student handouts).
• Refreshed several existing guides for recent product changes: Deep Research can now use recordings, podcasts, prior reports, and library files as sources; Quizzes and Exams now run as quick 10-question rounds with a recap; Folders now show how to spin up study guides, mind maps, infographics, and worksheets from a folder's contents; Flashcards covers the paid model selector; Uploading Content lists the new editable text and data file types.
• Refreshed the Video Modes guide to match the simpler three-mode picker (Standard, Math, Kids), and tightened the matching copy on the AI Video Lectures landing pages.
• Mistyped or short help URLs (for example /help/practice-exams or /help/notes) now resolve to the right article instead of a not-found page.

Bug Fixes

• Fixed AI video lectures and longer podcasts/recordings failing to play (showing an error) when the file was larger than about 32 MB.
• Fixed shared content links sometimes redirecting logged-out viewers to the login page.
• Fixed the flashcard study summary showing the full deck size for “Cards Reviewed” (and a wrong average time) when you studied a filtered set of a deck.
• Fixed flashcard answers occasionally rendering a stray red error instead of a dash in number ranges (for example 13–14%).
• Fixed sharing to WhatsApp or X from the share menu on mobile, which could silently fail to open.
• Fixed video chapters that could point past the end of a video, and stopped the post-study feedback prompt from reappearing after a page refresh.
• Fixed making flashcards from an uploaded image using up your free upload twice and then failing.
• Fixed the quiz upgrade prompt appearing on your last free question before you could answer it.
• Fixed the “Get Ultimate” button on the pricing page doing nothing for signed-in users.

Improvements

• The app now feels faster when opening pages, switching routes, searching your library, selecting page blocks, and following along with long video transcripts.
• Startup now shows your workspace sooner while larger library data continues loading in the background.
• Flip-mode flashcard study now reliably counts toward your streak and study stats, even if you close the tab mid-session.
• Added Mind Map to the “Create with AI” options shown after you add your sources.
• You can now turn product & marketing emails on or off with a single switch in Settings → Notifications, and every marketing email has a one-click unsubscribe. Your account and content-ready emails are unaffected.

New

• Editable library files. Text and data you create — notes, outlines, CSVs, JSON, code, diagrams, and more — can now be saved as real files in your library instead of one-off downloads. Each file gets its own page where you can preview it, edit it inline, download it, or restore it if you delete it.
• Chat with any file. Open a saved file and a side chat lets you ask questions about it, summarize it, or have it cleaned up — and edits the assistant makes are saved straight back to the file.
• Files you generate in chat and Deep Research can now be saved to your library in one click, and show up everywhere your other content does — Home, search, the sidebar, folders, and pinned — each with an icon that matches its type.
• Turn any saved file into anything. A saved file can now be used as a source for every AI tool — flashcards, quizzes, podcasts, video lectures, slides, study guides, worksheets, infographics, and mind maps — right from the file's own “Create with AI” menu or any create modal.

Improvements

• Every file type now has its own clear icon and color — spreadsheets, CSVs, JSON, code, diagrams, calendars, and more are instantly recognizable at a glance across your whole library.
• You can now upload many more file types directly (SVG, LaTeX, calendars, and a wide range of code and data files), and uploading a text file in any create flow saves it to your library automatically.
• Text and data files you attach in chat (notes, CSVs, code, and more) are now saved to your library as editable files and linked as sources — so they stay around and you can open, edit, or reuse them later, instead of being one-off attachments.
• Deep Research can now use far more of your own material as context: link your saved files, recordings, podcasts, and even previous research reports — not just pages and PDFs — so a report can build directly on your notes, lectures, and earlier work.
• Paid users can now choose the AI model for flashcard creation, with Gemini 3 Flash as the default and GPT 5.4 Mini available as an alternate.
• The whole app now respects your system Reduce Motion setting, with smoother, more consistent modals, buttons, menus, and page transitions throughout.
• Pages now load with content-shaped placeholders instead of a blank spinner — across Home, PDFs, videos, Deep Research, podcasts, recordings, and folders — and no longer flash a wrong “empty” state before your content appears.
• Clearer feedback in the moment: error messages stay long enough to read, can be dismissed, and offer a retry; the Create with AI panel now focuses the right field and tells you exactly what’s needed instead of a greyed-out button; the page editor shows a Saving/Saved indicator; and the video transcript follows along as the video plays.
• Refreshed the upgrade screen with a larger, easier-to-read layout, clearer plan benefits, and a tidier header.
• The “what’s included” hint on the upgrade screen now shows a clean checklist of everything your free AI creation credits cover.
• The Create with AI panel inside a folder now includes AI Study Guide, AI Infographic, and AI Mind Map, so you can build them straight from your folder's sources.

Bug Fixes

• Fixed a rare blank screen that could appear right after we ship an update — the app now reloads itself to the latest version instead of getting stuck.
• Fixed links and newly-created content that could occasionally fail to open — they now open reliably, with a copy-link fallback — plus the page editor’s “Type / for commands” hint not appearing and a flashcard flip/keyboard shortcut glitch.
• Fixed the Create shortcut buttons on Home looking washed-out when hovered in dark mode — they now highlight cleanly instead of showing a bright, mismatched border.
• AI video lectures now stream more reliably, especially larger generated videos on mobile and desktop browsers.
• Flashcard decks with math now load their formulas without occasionally showing a false “deck not found” error.
• Deep Research drafts no longer appear as stuck “Researching…” tasks before you submit a real research question.
• Bulk uploads now stop at your remaining upload slots instead of reading every extra file before showing the limit message.
• AI video lectures, infographics, mind maps, podcasts, and Deep Research now count as created only after the finished content is ready.
• Imports, source links, and research audio errors now show clearer recovery messages instead of misleading or blank error details.

New

• You can now create Mind Maps and Study Guides from Home. Both generate PDFs from your uploaded material, library sources, Drive files, links, or prompts.
• You can now create Worksheets and problem-set PDFs from Home, folders, uploads, or chat source actions.

Improvements

• Quiz sessions now run as quick rounds of 10 questions with a recap after each round — see your score, redo what you missed, or jump straight into the next round.
• Quiz questions for new decks are ready much sooner — the first batch appears while the rest keep generating in the background.
• When cards in a deck are due for review, the deck now shows a one-tap “ready to review” shortcut that starts a quiz on just those cards.
• Get an inline quiz question right while flipping cards, and Scholarly now offers to keep the momentum going in full quiz mode.
• When you reach a free-plan limit, Home now shows a clear banner telling you exactly what's maxed out (chat, uploads, creations, and more) with a one-tap way to upgrade for unlimited access.
• Deep Research now draws from your free AI creation credits — the same shared allowance that powers flashcards, podcasts, video lectures, slides, and more — instead of a separate research limit, so you can spend your free creations however you like.
• Locked AI models now show a lock badge for free users, and the upload window's Upload and Google Drive tabs now include icons with cleaner queued-item spacing.
• Home now keeps your recently opened generated PDFs visible in the Content list, even when they also appear in Jump back in.
• Your library looks consistent everywhere now — the All Content page and search use the same content rows as your Home page, with colored type icons and a cleaner, more readable layout.
• Mind Maps and Study Guides now appear in the create (+) menu and are included everywhere we list what your free AI creation credits can make.
• Clearer messaging when you reach your AI creation limit — it now explains the limit and the upgrade without implying your in-progress content won't finish.

Bug Fixes

• Math and formulas now display correctly in in-video quiz questions and the focus-mode question review (no more raw text like “d_model”).
• Deep Research no longer shows a “limit reached” message while a report is still running, and a fast double-tap can no longer start two reports at once.
• Completed flashcards, mind maps, study guides, and research no longer appear twice in your Home list once they’re ready.
• Generated PDFs like study guides and mind maps no longer show a stray “.pdf” in their title.
• Fixed the Study Guide theme picker mislabeling its themes as “slides”.

Mobile

• The menu drawer is friendlier one-handed — bigger New and Search buttons and larger Home and Chat tap areas, and long content, folder, and pinned-chat names now truncate cleanly instead of running under the actions button.
• Settings, Billing, and Connected Apps now open full screen on phones instead of a cramped floating box — with your account shown for context, the Upgrade card moved to the top, a collapsible usage breakdown, and a bigger Google Drive on/off switch.
• Chat works much better on phones: the web sources panel slides up as a full-width sheet, the drawing Canvas opens full screen with larger pen, eraser, and color tools, long chat titles no longer push the toolbar off-screen, and the Copy, Listen, and Retry buttons are bigger.
• Search opens full screen with results filling the screen, and every content row now shows its type and last-opened time (not just the title), with bigger pin and options buttons and long-press to open an item's menu.
• Deep Research switches between a report and its sources with a simple Report / Sources tab instead of a floating overlay, with bigger report action buttons (extras in a More menu), larger audio controls, a collapsible in-progress plan, and a full-screen Link Content window.
• Studying is smoother: tap the deck progress bar for your Mastered / Learning / Unseen breakdown and spaced review, the Filter & Sort, Analytics, and Create Exam windows open full screen, the exam question mix has quick percentage buttons, and the quiz Back / Next and flashcard create flow are easier to tap.
• Media is easier to play and read on phones: the PDF toolbar tucks zoom, download, and full screen into a menu, Slides mode adds tap-to-jump and a flip hint, the podcast play button is within reach with top-to-bottom chapters, video chapters peek to show you can swipe, and all playback controls are bigger.
• The create-with-AI and Import windows fit phones — tabs lay out as a clean grid, the Import button is pinned to the bottom, source buttons are bigger, and prompts use a shorter placeholder. Opening a folder that already has content now lands you straight in Chat.
• The page editor gives every block a tap-to-open menu (insert, change type, color, delete, and move up or down) so you can edit and reorder without a mouse; embedded images and PDFs no longer overflow the page; and Explore and generated charts use the full width with more readable layouts.
• Tooltips no longer flash and stick when you tap a button on a touchscreen, and tab switchers no longer overflow or clip their labels on small screens.

New

• You can now create AI content from Google Drive. Connect Drive in any create window, pick the exact files you want Scholarly to read, and turn docs, sheets, slides, PDFs, Word docs, PowerPoints, or text files into flashcards, quizzes, podcasts, AI video lectures, AI slides, AI infographics, and more — no downloading and re-uploading required.

Improvements

• Your connected Google Drive is easier to manage — settings now show at a glance when Drive access is active or needs reconnecting, with per-account disconnect.
• Adding Google Drive files to chat is now one click — the Drive picker opens straight from the content menu, and your picked files appear in the message box right away.
• You can now link Google Drive files directly in Deep Research — pick your docs, sheets, slides, or PDFs and Scholarly reads them as grounded sources for your report.
• Deep Research now opens to your past research — the question box sits up top and your previous reports are listed right below, ready to jump back into.
• The selected-text menu in chat is simpler now: it shows Ask, Copy, and Speak without the extra Cards action.
• Removed redundant AI shortcut buttons from file previews, exam results, focus-mode results, and the post-purchase resume flow so study screens stay focused on the primary action.

Bug Fixes

• Chat answers stream in live again — tool results, created content, and each part of a multi-step answer now appear as they arrive instead of all at once when the whole reply finishes.
• WPS Docs links now import from the actual document preview instead of the WPS loading page.
• Podcast generations with unusable prepared source text now fail quickly with clearer guidance instead of retrying the same empty script.
• Browser pop-up blocker messages for opening share windows and sources no longer count as app error alerts.
• File uploads, Google Drive connection denials, and flashcard creation failures now show clearer recovery paths instead of generic error messages.
• AI infographics now prepare their PDF study guide before completion, so summary generation no longer gets stuck on newly created infographic PDFs.
• Deleted research sessions now disappear from your library right away instead of lingering until you refresh.

Improvements

• You can now pick the video quality for AI Video Lectures — 480p for faster, smaller videos, 720p (the default), or crisp 1080p Full HD on paid plans.
• Improved text contrast and dark mode colors across the recording, video, podcast, and PDF pages.
• Text contrast across the app has been improved — buttons, menu items, tab labels, and table headers are now sharper and easier to read in both light and dark mode.
• Improved text contrast and card borders throughout Settings, Billing, and the sign-in pages for better readability in both light and dark mode.
• Study mode and quiz question text now renders correctly in dark mode, with proper contrast across all question types.
• Content lists, folder panels, and the source picker now have better text contrast and more consistent borders, making items easier to read and distinguish.
• Improved text contrast in the research report viewer, page editor, and website builder for better readability in both light and dark modes.

Bug Fixes

• Uploading a link that looks like a PDF but downloads a web page now shows a clear invalid-PDF message instead of failing like a server error.

Improvements

• Quiz sessions now ease you in: they open with a few easier questions, start each new card with multiple choice or true/false, and only move to written answers once you’ve gotten that card right before.
• Exams now open with a short warm-up and place written-answer questions earlier, while you’re still fresh — no more hitting the hardest questions right at the end.
• Flashcard decks now always play in the order of your material — your very first deck no longer jumps to the hardest cards first.
• The Full Exam preset is now 30 questions in 30 minutes — long enough to be thorough, short enough to actually finish. You can still build a bigger exam with custom settings.
• AI Video Lectures are back on the free plan! Creating a video now simply uses one of your free AI creation credits, just like flashcards, podcasts, and slides.
• Scholarly Ultimate is now much more affordable in 17 countries, including India, the Philippines, Brazil, Indonesia, Mexico, Nigeria, and Vietnam — prices are now adjusted to local purchasing power instead of a straight currency conversion.
• Failed or stuck generations now explain what happened more clearly, return any eligible AI creation credit, and follow up by email when a credit is restored.
• Dark mode is easier to find: there’s now a dedicated Theme tab in Settings where you can switch between Light, Dark, and System.
• You can now create an AI Infographic straight from the “Your content is ready” screen after uploading, alongside flashcards, quizzes, podcasts, and slides.
• The video style picker in the AI Video Lecture creator has a cleaner layout — Standard is featured as the default, with Math, Storybook, Doodle, and Kids below it.

Bug Fixes

• Reviewing a specific set of cards (like “redo missed”) now finishes when you’ve been through those cards, instead of quietly continuing with the rest of the deck.
• Practice rounds started after finishing a quiz session (such as “Redo missed” or “Review cards”) now count toward your study history and stats.
• Submitting an exam with unanswered questions is now noticeably faster — blank answers are marked as unanswered instantly instead of being sent for grading.
• Multiple-choice answer options are now shuffled evenly, so the correct answer no longer appears in some positions more often than others.
• AI Video Lectures now avoid a rendering path that could leave new lectures generating for far too long.
• The PDF viewer’s Flashcards tab now shows live progress when a deck is being generated from that PDF, instead of offering to create another one.
• Email sign-in now sends already logged-in users back into their account instead of leaving them on the recovery form, and missing-account messages include a direct support path for billing issues.
• Fixed PDF pages sometimes rendering partially cut off (for example a table missing its lower half) after zooming in or out — the viewer now always finishes drawing the page at the latest zoom level.
• Practice quizzes, exams, and answer checking now focus more reliably on understanding your material instead of trivia or unverifiable guesses.

New

• AI Infographics: turn your PDFs or a study prompt into a beautiful one-page visual summary. Pick from five visual styles — Editorial, Bold & Vibrant, Sketchnote, Minimal, or Freestyle — add custom instructions, choose the output language, select your AI model (paid plans), and get a poster-style PDF you can study, download, or share. Find it under "AI Infographic" on your home screen.

Improvements

• AI Video Lectures are far more reliable now — longer and more detailed lectures no longer get cut off partway through generating, and if a video hits a temporary snag it now recovers and finishes on its own instead of failing.
• Diagrams the AI makes for you — flowcharts, timelines, mind maps, and more — now open in a full viewer instead of being squeezed into the chat. Click the diagram to open it on the side, then zoom and pan to read it, switch to the source, or download it as an image. Your diagrams are also saved to your library.
• AI Video Lectures generate faster. The lecture builder now reads through your sources more efficiently and only takes a closer look at pages with diagrams, figures, or equations that it needs to see.

Bug Fixes

• Fixed a bug where editing a larger flashcard deck — deleting several cards, or refining a deck with AI — could fail to save and appear to revert or disappear after a reload. Your edits, deletions, and AI-refined decks now save reliably.
• AI page edits and saved answers are more reliable, including answers that contain diagrams, code snippets, or sections the AI removes for you.
• Submitting a timed exam after the timer ran out (for example when your laptop went to sleep) no longer throws away your answers — the exam is graded with your time capped at the limit.
• Asking the chat assistant to make flashcards from an image attachment no longer fails after starting — the assistant now reads the image and builds the deck from its content.
• AI Video Lecture styles, Deep Research starts, study answer checking, onboarding, and complex PDF processing are more resilient.

Improvements

• The app should feel noticeably snappier across Home, chat, folders, and content pages — faster navigation, smoother AI streaming, and quicker loading when you open your library.
• Dropping multiple files into a folder now uploads them in parallel instead of one at a time.
• Blog posts load instantly when you navigate between articles instead of freezing on the previous page.
• The Home page got a cleaner, calmer redesign. Quick-create tiles sit right up top, a new Jump back in section resurfaces the cards you have due for review and your most recent work so you can pick up where you left off, and every piece of content now has a colorful icon so you can spot its type at a glance.
• Your library looks consistent everywhere now — the content pickers in the create menus and the “Select Content” picker in chat match the new Home, with the same colorful type icons and cleaner rows.
• The Help Center got a refresh. Guides for AI Video Lectures, Video Modes, PDF flashcards, Plans & Limits, Quizzes, and the AI Assistant are updated to match the latest product — including the new Auto and Doodle video styles, the higher 32-page PDF-to-flashcards limit on Free, the post-session quiz review, and downloading your finished video as an MP4.
• AI Video Lectures are easier to understand before you generate: the landing page shows all teaching styles, the upload flow sets clearer expectations, and every lecture now covers the full source instead of asking you to choose a length.
• Practice quiz and exam questions now focus on understanding concepts and relationships rather than memorizing dates, years, or specific percentages.

Bug Fixes

• Images attached to Research requests are now used correctly, so reports stay grounded in the material you uploaded.
• The AI Assistant no longer leaves stray code-like text under interactive widgets.
• Math and chemistry now display properly when you review quiz or exam answers, instead of occasionally showing raw formatting.
• Pressing Submit on an exam before answering any questions now asks you to confirm, instead of appearing to do nothing.
• Making flashcards from a video we can’t transcribe — one with no spoken audio, or one too large to transcribe — now tells you exactly why and points you to a usable source, instead of repeatedly failing with a confusing “needs a transcript” message no matter how many times you tried. If a video is simply still being transcribed, you’ll now be told to give it a moment and try again.
• Downloading an image now saves reliably instead of being caught by your browser’s pop-up blocker — the download no longer opens a blank tab that can be blocked.
• Home page row actions now work correctly: pinning content and opening the three-dot menu no longer opens the item, and recent chats now have a Rename action from their own menu.
• AI Video Lectures no longer occasionally finish as a too-short, silent, or incomplete video. We now catch a broken render and retry it — refunding the attempt if it can’t recover — instead of delivering a stub.
• Auto-generated flashcard deck titles now keep apostrophes and hyphens — you’ll see “Newton’s Laws of Motion” instead of “Newtons Laws of Motion.”

Improvements

• AI video lectures now default to a new Auto mode that reads your material and picks the best style for it — so you get a great-looking video without choosing one yourself. You can still pick a style by hand.
• Added a new Doodle video style — hand-drawn sketch notes that build up on screen as the narrator explains, great for big-picture ideas, processes, and trade-offs.
• When you ask the assistant about something time-sensitive — like today’s weather, current prices, or live scores — it can now fetch the latest information from the web instead of relying on cached results.

Bug Fixes

• You can now review your quiz answers. When you finish a study session — or reach your daily quiz limit — you’ll see each question you answered in one list, marked correct or incorrect, with your answer and the right answer side by side.
• Newly created flashcards, quizzes, summaries, slides, and AI videos now open more reliably without false “couldn’t open” messages.
• Starting a new chat now updates the address bar cleanly instead of showing a false URL error.
• Fixed a crash that could happen while taking a flashcard exam.
• Your recent chats now appear on Home instead of Home showing an empty “upload your materials” screen when a chat was all you’d created.
• Uploading a corrupted or unreadable PDF now shows a clear message instead of a generic failure.
• AI video lectures and slides built from scanned or photographed notes (e.g. CamScanner PDFs) now use your actual pages instead of occasionally generating generic content from the web.
• AI chat no longer says it can't find your uploaded PDF (or other content) when you ask it to use “the pdf” — it now correctly looks up content by type.
• Finished Deep Research reports no longer stay stuck showing “Researching…” on Home — a completed report now flips to ready right away so you can open it.

Improvements

• Refreshed the Home page with a calmer, more focused layout — one clear place to ask AI or add a source, your study reviews and recent work brought to the top, and much less on-screen clutter.
• Free accounts can now turn up to 32 pages of a PDF into flashcards at once, up from 8 — so longer lecture decks and textbook chapters work without upgrading.

Bug Fixes

• Switching between study modes, video/PDF views, and Explore filters is instant again instead of reloading the page.
• Video references in AI chat now appear as clickable badges that jump to the exact moment in the video.
• When you ask the assistant to save an answer as a page, tables save properly and source references stay out of the final page.
• Flashcard decks created from website links now open reliably as soon as they finish generating, instead of occasionally showing that the deck could not be found.
• Flashcards now generate more reliably from image-only, low-text, or long files, with clearer guidance when a source genuinely needs to be changed.
• The “Prompt AI” tab now generates a flashcard deck from a topic you type, even without uploading a file, instead of failing.
• Study questions, deck chat, and PDF chat now stay grounded in the actual material you are reviewing.
• AI video lecture captions and the transcript no longer run past the end of the video, so clicking a transcript line always jumps to the right place.

Improvements

• AI video lectures are now fully usable on a phone — drag the progress bar to scrub, larger tap-friendly playback controls, and proper full-screen on iPhone.
• The Create with AI flow is smoother on phones — action buttons fit on screen, queued files and sources are easier to remove, and you can browse through folders without the path running off the edge.
• AI Chat is easier on mobile — a tidier input toolbar, a bigger button for attaching your own sources, and the model picker (including premium models) now opens with a tap.
• Studying on your phone is more comfortable — Focus Mode hints fit the screen and the next/previous buttons are easier to tap.
• AI video lectures can now finish at the length that best fits your material instead of being padded to hit a hidden minimum duration.
• Visual polish across study, upload, chat, and post-upgrade screens — calmer, more consistent icons and text, and tidier spacing in both light and dark mode.

Bug Fixes

• File uploads are more consistent across Scholarly: Word, PowerPoint, PDF, image, audio, video, text, and code files are recognized correctly in more places.
• After upgrading, blocked files are re-checked automatically in creators like flashcards, quizzes, podcasts, slides, and AI Video Lectures, so you do not need to close and reopen the window.
• If something is created successfully but cannot open automatically, Scholarly now gives you a Home fallback instead of showing a misleading creation or delete failure.
• Upload errors are clearer for protected PDFs, unsupported Office files, oversized files, and files that need to be selected again.
• Long-running creations are cleaned up more reliably, with clearer status and credit handling when something cannot finish.
• PDF study guides, summaries, citations, Help Center pages, AI podcasts, and AI video lectures are more reliable.
• Timed practice exams now keep their paused state if you leave and come back, so paused time no longer causes a false time-limit expiration.

Improvements

• You can now download your AI-generated video lectures — open the video and choose “Download video” from the menu in the top-right.
• Word and PowerPoint files now convert more reliably — if the main converter is temporarily unavailable, the app automatically retries with a second method instead of failing.
• Improved light mode contrast across the app — text, cards, and interactive elements are now sharper and easier to read.
• Consistent design polish across all dashboard pages for a cleaner, more cohesive experience.
• Improved text readability in the flashcard creator, customization panel, exam screens, and study selectors — descriptions and helper text are now noticeably clearer in light mode.
• Fixed several dark mode rendering issues in the sidebar and shared UI components — headings, icons, and overlays now display correctly in both light and dark themes.

Bug Fixes

• PowerPoint uploads now respect the same file-size limit shown for your plan, so paid users are no longer blocked by a lower hidden conversion limit.
• Fixed a rare case where the AI assistant could finish a reply with a blank message — it now gives clear next steps, like rephrasing or breaking the request into smaller parts.
• Uploading a PDF for flashcards now retries automatically when the document service is briefly unavailable, instead of failing with a misleading “no study material” message — and explains clearly when a file genuinely can’t be read.
• Word and PowerPoint files that can’t be converted now show a clear message naming the file and the likely cause (corrupted, password-protected, or an unsupported format), instead of a vague error.
• Flashcard decks no longer include duplicate cards that repeat the same question.

Improvements

• Spot a bug, a rough edge, or something you wish Scholarly did? Just mention it to the chat assistant — even in passing — and it’ll log your feedback for our team and confirm it right there in the chat, so it actually reaches us without you leaving what you were doing.
• New for AI Video Lectures: tap “Plan with Chat” to talk it through with the assistant first — it asks what to cover (specific chapters, pages, or sections), how long and how deep to go, and what to emphasize, then builds the video to your spec instead of you guessing at the settings.
• The video lecture instructions box now shows tap-to-add examples like “Only cover Chapter 1”, “Focus on pages 10–20”, and “Skip the introduction”, so it’s clearer how much you can tailor your video.
• Videos now open instantly and show a loading spinner while the video loads, instead of a blank screen that could feel stuck on a slow connection.
• The Explore and Search pages now open immediately and show a spinner while results load, rather than pausing on the previous page before they appear.
• When a new version of Scholarly ships, you now get a dismissible “new version available” banner and choose when to refresh, instead of the page sometimes reloading on its own in the middle of what you were doing.
• Flashcard decks open faster — highlights are now generated only when you actually open the Highlights tab.

Bug Fixes

• Fixed an issue where the Explore search page could get stuck on a loading spinner instead of showing your results.
• Fixed flashcard generation incorrectly reporting “no study material was provided” when your uploaded files couldn’t be loaded — it now clearly asks you to re-upload them.
• Chat now explains when a conversation has grown too large to send and how to continue, instead of showing a generic error.
• Fixed a rare crash in the notes editor while checking writing suggestions.
• Fixed AI image edits occasionally failing when based on an uploaded reference image.
• Fixed the “How was your voice session?” prompt appearing even when you only cancelled the microphone-permission popup.
• Flashcard highlights now retry automatically after a temporary generation hiccup instead of staying stuck.
• Adding YouTube or website links in the flashcard creator no longer counts against your file-upload limit, so the Create button stays available.
• Fixed the upload window getting stuck on the action screen after hitting your upload limit — you’re now taken back to the file step with a clear upgrade path instead of re-tripping the limit on every click.
• Fixed files that were too large staying blocked in the flashcard creator even after upgrading your plan in the same window — they’re now re-checked automatically.
• Fixed the “What’s New” popup appearing immediately after you closed the upgrade dialog.
• Fixed the feedback prompt reappearing on every visit even when you’d already seen and dismissed it.
• Fixed retrying a failed upload sometimes creating a duplicate copy of a file that had already uploaded successfully.
• When a PDF has more pages than your plan allows, flashcard creation now tells you exactly how many pages it could cover and how to unlock the rest, instead of a vague “content too large” error or a silently shortened deck.

Improvements

• The app should feel snappier across chat, your library, folders, and search — especially while AI responses are streaming and when you have a large collection of sources.
• Long study pages with many blocks load faster — blocks outside the viewport stay lightweight until you scroll to them.
• Chat messages with math and code render more smoothly, with less stutter while the assistant is still typing.
• Website and PDF links now work in more create windows, including AI Video Lectures and AI Slides, with clearer guidance when a page requires sign-in or human verification.
• After adding website content, AI Video Lecture is now easier to start from the Add Content window, while quizzes live under More with the other secondary actions.
• The setup screens in the "Create with AI" windows (adding content, flashcards, quizzes, slides, podcasts, and video lectures) now share one consistent layout — the heading stays at the top and the Back and Create buttons stay pinned to the bottom, instead of the buttons drifting down the page.

Bug Fixes

• When a research report’s audio can’t be generated, the report now shows a short explanation instead of leaving a blank space where the player would be.
• Video transcription now tells you when a file is over the 25 MB limit, instead of a generic “failed to extract transcript” message.
• Fixed the upload window getting stuck after an upload failed — if your file couldn’t be added, you’re taken back to pick a different one instead of being unable to continue.
• Fixed a rare error that could interrupt starting a study tool from one of our tool pages.
• Deep Research that stalls during generation now ends with a clear message instead of getting stuck “In Progress” indefinitely.
• Improved the reliability of file uploads when the app is handling a lot of uploads at once.
• Website links are now rechecked before we turn them into PDFs, so protected or verification pages do not slip through from a recent browser cache.

AI Video Lectures

• A brand-new player for AI video lectures and uploaded videos — with playback speed, a volume slider, one-tap captions, fullscreen, chapter markers on the timeline, and keyboard shortcuts.
• Narration now sounds like a real teacher: clearer, studio-quality audio at a brisker, more natural pace, with better emphasis and timing.
• The interactive questions that pop up while you watch are smarter and better timed — they now appear right after a concept is explained (instead of before it), are spread across the whole video, and test real understanding instead of trivia.

Help Center refresh

• New step-by-step guides for connecting Google Drive & Calendar, requesting access to private shared links, and using Scholarly on your phone — three flows that were missing a clear write-up.
• The Sharing, AI Slides, Voice Mode, and Video Lectures guides have been refreshed to match the recently shipped Share window, AI Slides visual themes, voice mic primer, and non-intrusive video questions.

Improvements

• The PDF Study Guide now has copy and print buttons so you can take your generated notes anywhere.
• Connected Apps settings now update Google Workspace actions with live status so Connect and Disconnect feel more responsive.
• AI podcasts and the read-aloud for research reports now play in crisp, studio-quality audio.

Bug Fixes

• Fixed stale app versions after updates showing a 500 error on video and other app pages; Scholarly now reloads onto the latest version instead of getting stuck.
• Fixed generated AI podcasts and video lectures getting stuck on “Loading” and never playing — they now stream and play correctly in the app.
• Research that could get stuck on “In Progress” forever now ends with a clear message when a report grows too large to save.
• Clearer, more accurate messages when flashcard, video, podcast, slides, or recording generation fails — we no longer blame your source material for what was actually a timeout or a temporary service hiccup.
• Flashcards that take too long now say so (and retry automatically) instead of suggesting you “use different files,” and the “no cards” message now explains the real causes and lets you try again.
• Recordings with no audible speech now say exactly that, instead of warning that the file may be corrupted.
• Fixed an issue where a long Deep Research run could fail to save its report. Research that genuinely can't be completed now stops quickly with a clear message instead of silently retrying.
• Fixed the date shown in AI chat being incorrect for some time zones.
• Fixed a rare error that could interrupt loading the model picker on chat and video pages.
• Fixed a crash on the PDF to Flashcards tool page caused by unusual characters in sample cards.
• Oversized images in chat now show a clear “over the 5 MB limit” message right away (with a prompt to resize), instead of a slow generic error.
• When a file upload fails to convert (e.g. an unsupported Office document), the error now stays pinned to that file instead of flashing by in a disappearing toast.

Improvements

• Simplified the Pricing page to a clear Free vs Ultimate comparison, removing a confusing duplicate column.
• The Create with AI panel inside folders can now collapse into an icon-only rail on desktop, and Scholarly remembers your choice.
• Background music in AI video lectures is now much softer — a faint bed under the narration instead of competing with the voice.

Bug Fixes

• Generated images in AI chat now open full-screen when clicked, so diagrams and visuals are easier to inspect across every chat.
• Fixed a brief flicker on the Pricing and Website Summarizer pages when they first loaded.
• Fixed an error that could stop the Explore search page from loading.
• Fixed the Help Center showing an error for visitors browsing the Japanese version of the site.
• Your finished AI Research reports no longer show a misleading "daily limit reached" message — the limit now only applies when you start or continue a research, with a clear upgrade option.

Request access to shared links

• Open a private link someone shared with you — a flashcard deck, PDF, video, podcast, recording, research session, or chat — and you can now ask the owner for access right on the page instead of hitting a dead end. The owner gets an email and can approve in one tap, and the content opens for you as soon as they do.