Back

NoBrainner Docs

Docs

Format guides for imports, video practice captions, and the technical docs we add later for API and integration work.

Remote MCP Server

Model Context Protocol (JSON-RPC 2.0)

Setup Guide Technical Specs →

Connect your NoBrainner spaced repetition decks, study stats, daily streaks, and weak card alerts directly to AI assistants like ChatGPT, Claude, Gemini, Cursor, Windsurf, and VS Code over Streamable HTTP.

Deck Creation

Deck import docs

Use these rules before uploading, pasting, or manually building a deck. They match the limits used by the app so users know what to fix before an import fails.

Safe limits

  • Paste and manual deck creation support up to 500 cards per deck. Split larger lists into smaller decks.
  • Free Anki .apkg limit: up to 100 MB per file, 1 upload/day, 5 uploads/month, 250 MB/month uploaded, and 500 MB stored media.
  • Pro Anki .apkg limit: up to 750 MB per file, 5 uploads/day, 50 uploads/month, 5 GB/month uploaded, and 5 GB stored media.
  • Anki decks also stop at 50,000 cards after parsing.
  • Quizlet PDF uploads support up to 20 MB. Quizlet DOCX uploads support up to 10 MB.
  • Manual image uploads support image files up to 8 MB before compression. Images are resized and saved as WebP.
  • Manual audio uploads support audio files up to 12 MB.
  • Bulk image search reviews up to 200 cards for free users and 500 cards for Pro users at a time.

File import steps

  • Step 1: choose the deck language, support language, deck type, and visibility before selecting a file.
  • Step 2: upload an .apkg, Quizlet PDF, or Quizlet DOCX. The file picker stays locked until the required deck settings are complete.
  • Step 3: follow upload and cloud-processing progress. You can leave after the upload completes; failed or canceled work is cleaned up automatically.

What works best

  • Best results come from clean front/back pairs, one idea per card, and the correct deck/support language chosen before import.
  • NoBrainner can read common Anki, Quizlet PDF/DOCX, pasted text, and manual cards. It cannot reliably fix messy files with mixed headings, commentary, tables, and extra notes without review.
  • Long sentence decks work, but image search is weaker when a card is a full sentence instead of a concrete object or action.
  • Very large Anki decks can process in the cloud, but browser upload, media size, and provider limits still matter.
  • Review extracted cards before creating when the source came from ChatGPT, Quizlet export text, or a scanned/odd PDF.

Paste formatting

  • Use one card per line.
  • Separate front and back with a semicolon, colon, or tab.
  • Use /n inside a card when one side needs a line break.
  • Do not add numbers, bullets, headings, or extra ChatGPT explanation lines.
  • Choose the deck language and support language before creating so shared pages show the right direction.
Paste example
I need more speaking practice.; Eu preciso praticar mais conversação.
Take care of yourself.; Se cuida.
Front line 1/nFront line 2; Back line 1/nBack line 2

Anki uploads

  • Upload Anki Deck Package files with the .apkg extension.
  • Keep scheduling information checked if you want NoBrainner to migrate due dates, intervals, reps, and lapses as closely as possible.
  • Keep media checked if you want images and audio. Without media, cards can import without sounds or pictures.
  • Deck presets are optional because NoBrainner uses its own study settings after import.
  • Leave old Anki support off unless you specifically need it; it usually makes the file larger and slower.

Quizlet PDF / DOCX

  • Quizlet PDF import expects a table-style export with terms and definitions.
  • Quizlet DOCX is the most phone-friendly option when the app offers a download.
  • If Quizlet only gives copied text, paste it into Create Deck and review the extracted cards before creating.
  • After PDF or DOCX extraction, NoBrainner sends the cards to the paste flow so you can choose audio, images, language, and sharing settings.

Manual cards and media

  • Manual decks and Manage Cards can use uploaded images or image search.
  • Uploaded images are compressed before saving so they sync across devices without keeping huge originals.
  • Image search works best with a short keyword or phrase. Long sentences are simplified before provider search when optimization is available.
  • Audio files should be real audio files such as MP3, M4A, OGG, WAV, or WebM.
  • Media belongs to the user copy of the deck. Editing your copy does not rewrite another user copy from a shared deck.

Sharing and duplicate safety

  • Public decks can appear in Shared after the deck language, support language, type, and visibility are saved.
  • Private decks stay out of the public catalog.
  • Shared links import a separate copy for the receiving user. Study progress and Anki scheduling from the owner are not copied as someone else's progress.
  • If a shared deck is already in the library, NoBrainner opens the existing copy instead of creating a duplicate.

Video Practice

How Video Practice works

Video Practice is a listening workspace for YouTube videos and timestamped captions. Browse or create practices, fix the transcript, and turn any useful lines into flashcard-style video clip decks.

  1. 1

    Start in the catalog

    Open Video Practice to see Continue watching, your categories, and Community videos. “See all” opens a dedicated page for the selected source or category, so it does not mix unrelated videos.

  2. 2

    Search and filter without loading everything

    Search titles, categories, or video IDs, then filter by spoken language and study language. Community results arrive in small pages; use Load more when you want more, instead of fetching thousands at once.

  3. 3

    Preview community videos first

    Clicking a Community video opens its practice preview. Watch and inspect the captions before choosing Add to my videos. Adding is explicit; a preview click never imports it automatically.

  4. 4

    Create or import a practice

    Paste a YouTube link and choose the video and study languages. Load captions from YouTube, upload SRT/VTT files, paste timestamped text, or use the available transcription fallback when captions cannot be read.

  5. 5

    Repair before studying

    Review the line list and translation, then use Video Settings or Studio to rename the practice, edit captions, replace translations, adjust timing, and choose private or public visibility.

  6. 6

    Practice line by line

    The player uses NoBrainner controls for previous/next line, play/pause, replay, speed, translation visibility, transcript search, and timing preview. Fun mode pauses normal navigation and asks you to rebuild each caption from scrambled words.

  7. 7

    Make a deck from useful lines

    From a saved practice, choose the current line, selected lines, or the entire video in Create a clip deck. Preview a card before creating it.

  8. 8

    Review clips like regular flashcards

    Video clip decks appear with your normal decks on the dashboard and keep the usual deck customization, folders, preview, settings, and review flow. They add only a Video Clips badge; cards use FSRS and the regular Again/Hard/Good/Easy XP values.

Caption sources and formats

  • YouTube captions may be creator-made or automatic. Availability depends on the video, language track, and YouTube access rules.
  • If YouTube captions are unavailable, upload .srt/.vtt, paste timestamped text, use the Chrome extension, or use an available automated transcription option.
  • You can add a translated caption file or translated timestamped text. Matching timestamps and order produce the most reliable pairing.
  • Long or broken captions can be edited in the library and cleaned with the available AI tools, while timing repairs stay separate from wording repairs.
  • If the automatic caption button fails, use the Chrome extension on a computer: play the part you want, stop the capture, then follow its progress on Video Practice while the captions are prepared. Review it here, save it to your library, and create video flashcards.

Library, community, and organization

  • Your saved video practices are stored in your account and cached locally for faster reopening; the library keeps a small working set rather than pulling an unlimited collection into the browser.
  • Create categories, rename or delete categories, and move saved videos between categories from Organize.
  • Public practices can appear in Community videos. Shared copies are private to the receiving learner until they choose to share them.
  • Remove a saved practice from its trash button or video settings. Community previews remain available for browsing until you add them.

What happens after saving

  • Open a saved video to continue from its remembered line and playback position.
  • Use the transcript search and line controls to focus on one phrase, repeat it, slow it down, or hide the translation before listening again.
  • Fun mode is optional. Close it to return to normal line navigation; while it is open, finish the current scrambled-caption answer before moving on.
  • Video settings and Studio work on your saved copy, so caption or visibility changes do not rewrite another learner’s copy.

Video clip decks and FSRS

  • Each clip card keeps the video, exact timestamp range, caption in the video language, and your translation together.
  • On the front, play the clip and control its speed. Flip to the back for the translation and Again, Hard, Good, or Easy grading.
  • Swipe left/right on the back on phones when motion is enabled; reduced-motion users get the regular grading buttons instead.
  • XP is awarded using the same grade values as regular decks and is counted once per card per study day, including safe retry sync after a connection failure.
Learn how to use the extension

Video Practice

Caption import formats

Use one of these formats when YouTube captions are missing, blocked, or wrong. The first caption source should match the video audio. Translation captions are optional.

SRT / VTT subtitle files

  • Accepted extensions: .srt and .vtt.
  • Each cue needs a start --> end timing line and text below it; separate cues with a blank line. SRT sequence numbers and the WEBVTT header are supported.
  • SRT commonly uses commas and VTT commonly uses dots, but the importer accepts either: 00:01:23,000 --> 00:01:25.000.
  • Upload the original-language file first. If you have a translated file, upload it as the translation file.
SRT example
1
00:00:05,000 --> 00:00:07,000
You're getting to be a big boy.

2
00:00:07,500 --> 00:00:10,000
I'm just a kid who's four.
VTT example
WEBVTT

00:00:05.000 --> 00:00:07.000
You're getting to be a big boy.

00:00:07.500 --> 00:00:10.000
I'm just a kid who's four.

Pasted timestamped text

  • Accepted timestamps: m:ss, mm:ss, h:mm:ss, and optional milliseconds with . or ,.
  • The timestamp can be on its own line or at the start of the same line; text until the next timestamp becomes that caption.
  • Blank lines between captions are optional.
  • Bracketed timestamps like [04:11] also work.
  • If you paste a translation too, keep the same timestamps and line order so sync stays predictable.
Timestamp on its own line
0:05
You're getting to be a big boy.
0:07
I'm just a kid who's four.
Timestamp and text on the same line
00:00:02 Damn Rolos.
00:00:05 - Hey, you're back. - Hey.
00:00:07 How was your conference?
00:00:08 It was terrible, I fought with..

Third-party transcript sites can help, but they are not part of NoBrainner. They may show ads, track usage, or change formats. If you use one, turn on its timestamp option before copying text back here.

Future docs

This page is the home for future API notes, integration guides, import rules, and troubleshooting docs. New docs can be added as separate sections without turning one file into a monster.