Scholarly August 2026 Updates Recap

August was less about one headline feature and more about Scholarly becoming a genuine workspace: Projects grew from a simple container into a shared home for your sources, chats, and creations; Deep Research and AI Meeting Notes were rebuilt from the ground up; and Home, Explore, and the pricing pages all got cleaner, calmer redesigns. Alongside that, Scholarly's interface reached Japanese users in full, and local pricing rolled out to a dozen more countries. Here's the full rundown.

---

Major Features

Projects: One Workspace for Everything You're Working On

Projects launched as a dedicated workspace that keeps the sources, chats, and AI creations for one goal together, with project instructions that every chat inside it follows. Any folder can become a project in one click, bringing its contents along, and you can move existing chats and content into a project (or back out) from the row menus in your library, sidebar, and chat list. Project memory shows what Scholarly has picked up while working inside a project so you can retire anything out of date, and every creation modal now shows a "Saving to" chip so sources and finished work land in the same project automatically.

Through the rest of the month, project chat kept gaining real work to do: it can rename, archive, restore, and move your pages, files, and folders on request, create new folders, and — when a conversation turns into ongoing work — offer to set up a project for it. It can build something for a time you name ("have this ready before my exam Friday") and file it on Home as a scheduled task. It can also propose edits to your pages, flashcard decks, and files directly: every edit arrives as an approval card you accept or deny, so nothing changes without your say-so.

By the end of August, shared projects became real collaborative workspaces. People you invite can open a shared project from their own Projects list, chat across its sources, edit the shared brief, and see who else is active, while keeping their own private chats and memories. Only the owner can add or remove content, manage members, or archive the project, and sharing gives collaborators a read-only view of everything in it — files, pages, and folders — while project chats stay private.

How to use it: Open Projects from the sidebar, name a project (or convert an existing folder), and add your sources. Chat inside it for answers grounded in that material, and anything you create there is filed back into the project automatically.

Deep Research, Rebuilt as a Real Research Workspace

Deep Research became a professional workspace with its own sidebar destination instead of a single long report. Scope a run to a Project, attach internal sources, choose the evidence mix, depth, and model, or start from a ready-made workflow — literature review, market landscape, policy analysis, or due diligence. Finished work opens with a concise verdict, confidence and contested-claim signals, and open questions, with the complete cited report in its own Research tab, sources in Evidence, and generated briefs, tables, charts, and working files in a full-screen file workspace you never have to leave.

A dedicated chat sits beside the research and already knows its reports, findings, and sources, so you can ask for a grounded summary, compare findings, or challenge a conclusion without leaving the report. It's also the one place to request follow-up research — a new report starts in the same session, with report history sitting beside the chat.

How to use it: Start Deep Research from Home or the sidebar, describe your objective, and pick at least one deliverable — a cited report, slides, a brief, a study guide, a table, or your own combination.

AI Meeting Notes: Watch Your Recording Transcribe Itself, Live

Recording got a full overhaul. AI Meeting Notes now stream the transcript as you speak, with a live status telling you whether Scholarly is listening, transcribing, or hearing nothing at all — and the moment you finish, the transcript is already there while the summary and action items are written. Capture itself moved into one small floating bar that follows you across the whole app, with elapsed time, a live level meter, and pause/finish always one click away, so you can open a source or start a chat without ending the recording. The recording page itself is calmer too: title, status, and date up top, then one card holding Transcript, Summary, Action items, and AI Creations.

Recording also got sturdier: live capture now detects when your computer sleeps, a tab is suspended, or the microphone drops mid-recording, restarts the mic where it can, and discloses on the finished notes exactly how much time went uncaptured instead of quietly delivering a partial recording.

How to use it: Start a recording from Home or the floating capture bar — it follows you anywhere in the app, and you can keep working while it listens.

One AI Document Creator for Every File Type

PDFs, reports, timelines, SOPs, spreadsheets, mind maps, worksheets, and infographics now all start from a single AI Document Creator, with one source picker and a document-type chooser instead of a dozen separate tools. Pick a preset type, or choose "Anything" and describe the document you need in your own words.

A Calmer Home, One Input for Every Tool, and a Rebuilt Explore

Home's input now carries the whole toolbox along its top — pick Chat, Image, Document, Deep Research, or Essay and describe what you want, or pick Study Guide, Slides, Mind Map, or Worksheet and drop a file straight onto it. What you type or drop is carried into that tool's setup instead of being asked for twice. Home also leads with a continue card that names your single next step, replacing the old welcome-back popup, and the whole dashboard got a design refresh: buttons, inputs, and panels now sit on subtle background steps instead of boxes and outlines, and every plan limit looks and reads the same wherever you meet it.

Explore, Scholarly's public library, was rebuilt as one coherent showcase instead of flattening every kind of work into the same card — Short Videos are vertical, Podcasts have episode artwork and waveforms, Flashcards show a real question, and Research shows its abstract and sources. It's fully available in English and Japanese, with clearer search and complete coverage for both.

How to use it: Type or drop straight onto the Home input to jump into any tool — no separate "create" step needed.

Video Lectures Get Color, Dark Mode, and a Free Companion Slide Deck

Standard AI Video Lectures now draw from a pool of hand-tuned color palettes — soft pastels, brights, and editorial looks alongside the classic cream-and-blue — so each new lecture gets its own look, and a Dark Mode toggle renders every slide on a matching dark theme. Every Standard lecture now includes a companion PDF slide deck by default, on every plan including Free; it opens in its own Slides tab, and Premium members can download it beside the video.

Podcasts Get Producer Notes, Related Resources, and a Video Mode

New podcasts now include timestamped Producer Notes that add researched context at the moments it matters — jump to the exact point, follow the cited links, or send a note into the podcast chat to explore further — plus a Related Resources section curating trustworthy sites, papers, and courses for continuing beyond the episode. The player itself now docks in a bar pinned to the bottom of the page, and podcasts with chapter slides gained a synchronized Video mode with a share-ready MP4.

Three New Standalone Tools

Text-to-Speech reads any document aloud in full and saves it as one MP3 in your library — pick from over 25 voices and a 0.75×–1.5× speed, with Scholarly quoting the exact cost before you start. Transcribe turns up to 60 minutes of audio, video, or a YouTube video into a clean .txt file or timestamped .vtt captions, running in the background and returning your credit automatically if it fails. And the AI Image Playground graduated out of beta — no more daily cap, each generation uses one AI credit per model, and images made or edited in regular AI Chat still don't use a creation at all.

How to use it: Find Text-to-Speech and Transcribe under More tools on Home, or from the creation actions on any file, page, or PDF.

Scheduled AI Creations and Images as First-Class Files

You can now schedule any one-time AI creation for a specific day and time instead of starting it immediately — scheduled work stays together on Home, or in a project's own Tasks view if it was created there. The same week, images (JPG, PNG, GIF, WebP, BMP, TIFF, HEIC, HEIF, AVIF) became first-class files you can upload anywhere you add sources: open them in the file viewer, rename, move, chat with them, or use them for Flashcards, Practice Quizzes, Podcasts, and Video Lectures.

Customization Profiles and a Simpler AI Credit System

Customization Profiles replaced separate Saved Instructions and Custom Designs with one reusable profile combining instructions, a preferred AI model, and visual references like logos or brand guides, applied from a button at the top of every Customize step. Existing saved instructions and designs moved into personal profiles automatically, team admins can publish shared Team profiles, and each Project can set a default profile for consistent chats and creations.

Alongside it, AI creation pricing simplified into one credit balance: standard models use 1 credit, and Frontier models like GPT 5.6 Sol and Claude Opus 5 use 3, replacing the separate Frontier Boost allowance.

Scholarly for Teams: Simpler Pricing, Guided Setup, Real Shared Workspaces

Teams pricing simplified to two seat plans — Premium for Teams and Laureate for Teams, self-serve from 1 to 29 seats — with every seat inheriting that plan's models and limits under one bill and no separate team credit wallet. Creating a team now walks through a clear guided flow: name the workspace, invite teammates by email or shareable link, choose a plan, then set seats, with a free trial that includes 3 seats so you can bring people in before paying. Members and admins can now leave a team themselves from the Team page, freeing their seat immediately.

Regional Pricing and a Rebuilt Pricing Page

Scholarly's pricing page was rebuilt with Personal and Business tabs so you can price a plan for yourself or per seat for your team without leaving the window, and it now matches the same plan cards used in-app. Local pricing rolled out broadly through the month — Premium now shows and charges matching local prices across a dozen more countries including Japan, Switzerland, South Korea, and the UAE, and Laureate added local-currency billing everywhere it's offered, so what you see is exactly what checkout charges.

Japanese Localization Reaches the Whole App

Japanese support expanded from a handful of pages to the entire signed-in product over the course of the month: Login and Home first, then Connections and core settings, then Library, Projects, and Chat, then every remaining corner — sign-up, Teams, sharing, study modes and exams, editors and exports, recordings, podcasts, video, Research, billing, and background task progress. Every email Scholarly sends — creation-ready notices, billing, renewals, team invites — now arrives in Japanese when your app language is Japanese, and with AI Language set to Automatic, podcasts, documents, essays, and research reports now follow your source material's language instead of defaulting to English. App and AI language are now separate controls, so changing your interface language never changes the language AI uses to answer or create. Signed-in URLs stay the same clean addresses throughout — only public marketing pages use a language-specific address.

Google Drive Connections, Done Right

Connections got its own place in the sidebar with per-account approval, default, reconnect, and disconnect controls. Access is read-only and per-file — Scholarly only sees what you share through Google's picker, including Docs and Slides, and Chat can search and read those files but never browse the rest of your Drive or edit anything.

---

Chat, Home, and Everyday Improvements

• A full voice picker replaced the cramped dropdown everywhere a voice is chosen — chat settings, Voice Mode, read-aloud, podcast guest voices, and video narration — with search and a short demo of any voice before you commit.
• Chat answers are shorter and lead with the answer, with less preamble and no closing recap, and now render in a lighter body weight that's easier to scan on longer replies.
• Command palette search (Cmd/Ctrl+K) now returns your project chats and conversations held over a PDF, page, recording, podcast, video, or folder, matching on what was actually said rather than titles alone.
• Chats can now reference earlier chats — type @ to find a previous conversation and bring its messages into the current discussion.
• The flashcard editor gained visible Undo and Redo controls, plus Cmd/Ctrl+Z shortcuts, for text, order, images, and added or deleted cards.
• Every flashcard deck's Sources tab is now AI & Sources — create a podcast, video lecture, or study guide straight from a deck's original material, and see everything else already made from it.
• You can close the tab while Chat is answering and reopen the conversation later without starting the response over.
• New accounts can buy one-time AI creation credit packs when a weekly allowance runs out, with clear math showing what packs cost against what a plan would include.

---

Mobile

• Explore, the blog, and file previews all read properly on a phone — titles no longer cut off, charts fit the screen, and comparison tables scroll sideways instead of losing columns.
• Dragging the podcast and recording progress bar now scrubs reliably instead of jumping to a chapter marker.
• Model pickers and other menus opened near the top or bottom of the screen now flip or slide into view instead of opening off-screen.
• A notification offering an action — like opening a deck the moment it finishes — now waits for you instead of disappearing after a couple of seconds.
• Focus mode results, exam answer choices, the image playground, and export dialogs were all fixed where they used to cut off or crowd out their controls on a small screen.

---

How to Use the Highlights

Start a Project

1. Open Projects from the sidebar and name a workspace, or convert an existing folder in one click.
2. Add your sources — files, links, pasted text, or Google Drive items.
3. Chat and create inside it; everything you make is filed back into the project automatically, and you can invite others to collaborate.

Run a Deep Research report

1. Open Deep Research from Home or the sidebar.
2. Describe your objective and choose evidence, depth, and at least one deliverable — a report, slides, a brief, or a table.
3. Review the verdict and confidence signals, then use the built-in chat to ask follow-up questions or start another report in the same session.

Capture a meeting or lecture live

1. Start a recording from Home — the floating capture bar follows you anywhere in the app.
2. Watch the transcript stream in as you speak.
3. Finish, and your summary and action items are ready moments later.

Build any document with AI Document Creator

1. Open AI Document Creator from Home or the More menu.
2. Add your sources.
3. Pick a document type — or choose "Anything" and describe what you need.

Schedule a creation for later

1. Start any one-time AI creation from a creation window or Chat.
2. Choose a date and time instead of Create now.
3. Find it on Home (or your project's Tasks view) — run it early anytime, or let it start on schedule.

Set up a reusable Customization Profile

1. Open the profile button at the top of any Customize step.
2. Combine instructions, a preferred model, and visual references like a logo or brand guide.
3. Apply it to any future creation — team admins can publish shared Team profiles for everyone.

---

Bug Fixes and Reliability

AI Chat

• Fixed the biggest one: the context budget for a chat turn had been outgrown by the instructions Scholarly sends with every message, so most turns were handed only your newest message and the reply just before it — a twelve-exchange conversation reached the model as three messages. The budget now leaves real room for history, so earlier corrections and context stay in view.
• Chat generations now recover from a provider safety stop by moving to a different model instead of ending permanently on the first attempt.
• If the file workspace behind a chat reply can't start, the assistant now says so once and answers with what it already has, instead of retrying the same step in a loop.
• Long research-style chat turns now get a reminder as they approach their step budget, so they write up what they found instead of finishing empty-handed.
• Fixed chat replies that could fail permanently when one AI model rejected a long conversation — they now move to another model instead of getting stuck.

AI Video Lectures and Podcasts

• Fixed lectures being handed to a retired model with no backup during automatic recovery; retired models are no longer offered where they've been withdrawn.
• Video narration now waits out a brief speech-service outage instead of failing the whole lecture, and keeps completed audio when a scene runs a few seconds long.
• Playing a long podcast, recording, or video no longer fails once it passes about 32 MB — files now stream in parts, so playback and seeking work the whole way through.
• Fixed podcast chapter navigation snapping back to the previous chapter instead of landing where you clicked.
• Two-speaker podcast requests now stay limited to exactly two voices throughout transcript and audio generation.

Flashcards, Slides, and Documents

• Flashcard decks now repair duplicate or missing internal card identifiers automatically whenever a deck is opened or edited, so a damaged deck no longer fails while preparing study questions.
• Generated Slides now use the whole canvas, keep tables in readable rows and columns, and no longer leave the bottom quarter of a slide empty.
• Fixed true/false exam and practice questions occasionally marking a correct restatement as false, and Chat, Slides, and Video Lectures now preserve exact signs, percentages, and identifiers from your source instead of normalizing them.

Uploads, Sources, and Search

• Empty (0-byte) files — an undownloaded cloud placeholder, a recording that never finished writing — are now caught immediately with a clear message instead of uploading and failing later.
• YouTube imports now ride out transcript-provider hiccups and outages instead of failing outright, and a video whose transcript is still generating is added to your library right away.
• Fixed search permanently falling back to a slower matching method for anyone served by a machine caught mid-deployment; it now recovers automatically once the fast index is ready.

Billing and Plans

• Upgrade prompts now open the plan screen that actually matches what you hit, instead of a generic or mismatched limit message.
• The Laureate upgrade card shows the prorated amount you'd actually pay today, with Stripe crediting time already paid on your current plan.
• Fixed switching an existing subscription to a different currency or plan sometimes failing with a server error on checkout.

Everyday Reliability

• Deleting your account now signs you out and confirms right away while removal finishes in the background, instead of appearing to fail during a longer deletion.
• Signing out or switching accounts in the same tab now fully clears the previous account's content from Home, the sidebar, and search.
• Fixed a translation-context error that could stop an uploaded source from reaching the next step, or make an upgrade button do nothing.
• Continued light-mode and dark-mode contrast passes across Settings, Billing history, and the cancellation screens.
• Faster loading and instant return-visits across Home, the Library, and Projects.

---

Looking Forward

August was the month Scholarly stopped being a set of separate tools and started feeling like one workspace: Projects grew into a real shared home for your work, Deep Research and AI Meeting Notes became genuinely professional-grade, and Home, Explore, and pricing all got calmer, more consistent redesigns. Regional pricing and Japanese localization also reached far more of the world this month. Expect Projects, Deep Research, and the unified Document Creator to keep expanding through September.

If you have any questions or feedback, reach out at hello@scholarly.so.