Education & Learning

Character.AI (Menlo Park; founded Nov 2021 by Noam Shazeer + Daniel de Freitas; Shazeer/de Freitas licensed to Google Aug 2024, Karandeep Anand CEO Jun 2025) = the YouTube of AI chatbots​ — 10M+ community characters, proprietary LLM (PipSqueak 2 free / DeepSqueak c.ai+), voice calls, group rooms, Imagine Gallery, AvatarFX, Stories mode. Not agentic, not API-exposed, not a productivity assistant. 

Citation Machine — Browser-Based Academic Citation & Bibliography Generator with Grammar/Plagiarism Upsell (Chegg-owned, Active, Not Shut Down)Paste a URL/ISBN/DOI or fill source fields → formatted in-text + reference-list entry in APA 7 / MLA 9 / Chicago 17 / 7,000+ styles; free tier ad-supported, premium grammar+plagiarism behind Chegg login.

Coconote — AI Note-Taker That Turns Lectures, PDFs & YouTube into Notes, Flashcards & Quizzes (Apple-Native, Free Tier)

Curipod — AI-Native Interactive Lesson Platform: Topic → 30-Second Live Lesson (Polls + Draw + Word Cloud + AI Writing Feedback, No Student Logins)

DeepL (Cologne, 2017; spun out of Linguee 2009; Index Ventures $2B val 2024) = the European-quality NMT specialist that beat Google Translate on fluency and now wraps itself in DeepL Write + DeepL Voice + Translation Flow — transformer NMT trained on Linguee bilingual pairs, 33 text languages / 100+ document languages, formal/informal toggle, glossary + style rules + translation memory. Not Google Translate (free/249 langs/shallower), not ChatGPT/Claude (context-aware but slower, no glossary/SSO pipeline), not Lokalise/Phrase (localization workflow layer on top).

Duolingo Max (Pittsburgh, 2011; Luis von Ahn + Severin Hacker; NASDAQ: DUOL, FY2025 rev >$1B) = the top paid tier of the 500M-user gamified language app, repositioned in 2023–2026 as an AI conversation tutor​ — Video Call with Lily (GPT-4o/5.x driven, adaptive spoken chat) + AI Roleplay (cafe/hotel/check-in scenarios) + Explain My Answer (moved to free tier Jan 2026) + Mistakes Hub, sitting on top of Super Duolingo (ad-free, unlimited hearts, offline). Not Babbel (structured human-written courses), not Pimsleur (audio-first pronunciation), not Speak (AI-speaking-first, no gamified tree), not ChatGPT (general, no spaced-repetition curriculum). Fits ai-productivity-data; pipeline_stage none (spans vocab-drill + grammar-explain + roleplay-speak + video-call-practice). Table-first, Claude Code skeleton. The 2026 trap: Jan-2026 Explain-My-Answer went free, so Max's paid edge collapsed to two features (Video Call + Roleplay) at ~$168/yr vs Super ~$96/yr​ — nearly 2x for speaking practice only, and Roleplay/Video Call limited to 7–8 languages.

Eduaide.AI (Annapolis, 2023; Derek Johnson + Thomas Hummel/Thompson, 80万+ educators, ~6 employees) = AI instructional-design workbench built by teachers​ — 100–110+ guided resource generators (Lesson Plan / Graphic Organizer / Game / Assessment categories) + Erasmus Smart Document Chat + Adapt & Differentiate + grade-level Evaluator + Feedback Bot, all grounded in retrieval/scaffolding/formative-assessment pedagogy. Not MagicSchool (broader ecosystem + student product), not Brisk (Chrome-side real-time assistant), not Diffit (text-leveling-only), not ChatGPT (chat, no standards/doc workflow). Fits ai-productivity-data; pipeline_stage none (objective → generate → differentiate → evaluate → export LMS). 2026 trap: Free = 15 generations/month (or daily limits on 81 tools) — pinch daily users into $5.99/mo fast; outputs fact-check and standards-alignment-claim still need teacher eyes; smaller ecosystem than MagicSchool for district adoption.

Explainpaper — Highlight-to-Explain AI Reader for Dense Research Papers (Context-Grounded Plain-English Fixes)

Grammarly = inline English writing co-pilot everywhere you type: keystroke → grammar/spelling/punctuation/clarity/conciseness/tone inline → Pro unlocks full-sentence rewrite, fluency, inclusive-language, citation consistency, plagiarism + AI-text detection, 2,000 AI prompts/mo → Docs editor + agent tray (Proofreader/Paraphraser/Reader Reactions/Citation Finder/Fact Checker/AI Grader/Humanizer) → Authorship panel logs writing provenance → Enterprise adds SAML/SCIM/DLP/BYOK/audit-logs/confidential-mode. Fits ai-productivity-data; pipeline_stage none (type→correct→rewrite→detect→export). For ESL writers, daily email/Slack/document drafters, students polishing essays, and teams wanting consistent tone without a styleguide workshop, it's the lowest-friction default; for long-form novel drafting, multilingual, or brand-voice enforcement at scale, ProWritingAid/QuillBot/LanguageTool/Writer win.

Khanmigo = patient AI tutor, not homework-answer machine: sign in (parent account for under-18s) → pick a Khan Academy lesson or just ask → Khanmigo asks guiding questions, breaks the problem into steps, pulls the relevant KA video/practice → student works it out → writing coach critiques essay structure without rewriting it → Debate Me argues both sides → coding sandbox (Python/JS/Java + 9 more, auto-grading, 2026 Khanmigo 2.0) → SAT/ACT/AP prep with personalized practice → teacher side: standards-aligned lesson plans, rubrics, exit tickets, quiz questions + answer keys, differentiation, student-work summaries, class insights dashboard. Fits ai-productivity-data; pipeline_stage none (ask→guide→practice→assess→report). For K-12 learners inside the KA ecosystem, homeschool families, and U.S. teachers drowning in prep, it's uniquely safe + cheap; for university/grad, non-KA curricula, or "just give me the answer," it's the wrong shape.

Learn.xyz = dead docs → addictive micro-courses: consumer opens iOS/Android → type "how does QE work" → Lumi generates 5-slide Curiosity + quiz + streak in ~10s, multi-language, social feed shows friends' topics → enterprise admin uploads 20-page PDF/handbook/deck → AI restructures into 6-8 micro-lessons with checks between chunks → edit in browser before publish → learners get mobile/desktop web app, gamified quizzes, monthly leaderboard with prizes → admin dashboard shows 89% completion (vendor-reported, Mar 2025 58 lessons/user/mo), compliance export, SCORM/LTI on Turn-key, US/EU data region, SOC 2 Type II, GDPR. Fits ai-productivity-data; pipeline_stage none (source→restructure→gamify→assign→complete→report). For HR/L&D rolling out policy/onboarding without rebuilding Slide decks, and for casual learners wanting Wikipedia-depth in Duolingo-format, it clicks; for university-credit courses, deep adaptive tutoring, or creator-monetize storefronts, it's the wrong shape.

Independent study tool (not Google/OpenAI subsidiary), ~5M+ users, 20+ languages, TikTok/Instagram student-driven growth. No acquisition, no sunset. Category ai-productivity-data, pipeline_stage=none. Edu/productivity lane — not in your Dev/Audio/Video 11-card set, but same card schema applies if you want it parked under ai-productivity-data with edu note.Official site: https://www.mindgrasp.ai​ Pricing: https://www.mindgrasp.ai/pricing Mindgrasp AI — AI Study Companion That…

QuillBot = the rewrite bench, not the draft desk: paste a paragraph into quillbot.com or the Chrome/Edge/Safari extension or Word add-in → Paraphraser rewrites with synonym slider + mode pick (Formal for memos, Academic for papers, Creative for blogs, Shorten/Expand for length control) in 23 langs → Grammar Checker fixes in 6 langs (EN US/UK/CA/AU, DE, FR, ES, PT-BR, NL) → Summarizer condenses 6,000 words (Premium) to bullets/paragraphs → Translator 40+ langs → Citation Generator (APA/MLA/Chicago/Harvard from URL/DOI) → AI Humanizer (machine text → "human score" + tone insights) + AI Detector (text + image) + Plagiarism Checker (25k words/mo Premium) → QuillBot Flow (doc editor, 100MB Premium) + AI Chat (20 free/∞ Premium) + AI Image (3 free/300 Premium) + PDF tools (ChatPDF/merge/split/convert). Fits ai-productivity-data; pipeline_stage none (paste→paraphrase→grammar→summarize→cite→humanize). For students/academics/ESL/agencies refactoring existing copy, it's the cheapest serious seat; for zero-draft generation, ChatGPT/Claude.

Quizlet = the "someone already made the deck" lane: open quizlet.com / iOS / Android / Win-mac-web → search "AP Bio photosynthesis" → 200 sets exist → clone one → hit Flashcards (flip), Learn (adaptive wrong-card resurfacing, capped free), Test (auto quiz), Match (timed pairing), Quizlet Live (class race) → paste your lecture notes or upload PDF → Magic Notes​ (OCR + LLM) spits flashcards + outline + essay prompts + Quick Summary + Memory Score → AI Practice Tests from your set (3/mo Plus, ∞ Unlimited) → Expert Solutions step-by-step (20/3 per mo Plus, ∞ Unlimited, EN textbooks only) → formerly Q-Chat Socratic tutor (dead Jun 2025), now Mar 2026 conversational study + Coconote audio/video ingest. Fits ai-productivity-data; pipeline_stage none (search/own set→magic notes→learn/test/match→live). For term-definition cramming (languages/anatomy/law vocab/certs) it's the lowest-friction seat; for true interval-based long-term retention, Anki+FSRS; for document synthesis, NotebookLM.

scholarcy (Scholarcy-Ltd London-2018 acquired-Oct-2024-by-Everway-formerly-Texthelp active-2026-09 not-shut-down 600k-users 170-countries 1M-summaries 100M-refs-extracted 99.94pct-uptime status-operational) = ai-academic-paper-summarizer-to-structured-flashcards not-tutor-not-discovery-engine: Flashcard-Summary KeyHighlights-Synopsis-Methods-Results Enhance-reading-level Spotlight Robo-Highlighter table-figure-to-Excel reference-extraction-open-access-links Dig-Deeper-QA cross-paper-comparison Literature-Matrix-100-summaries bibliography-Word Zotero-Mendeley-sync Google-Drive YouTube notes-highlights-annotations-collections searchable-Library Chrome-Edge-Firefox-ext Ref-Parser-API-JSON-from-225-mo; Free-Article-Summarizer $0 1-per-day Free-Account 10-summaries Library 4.99-mo Scholarcy-Plus 9.99-mo-7day-trial-unlimited-100flashcard-export Literature-Matrix Plus-annual 59.99-89.99-yr Institutional-from-8000-yr; starting_price=$0; category ai-productivity-data pipeline_stage none; competitors SciSpace Elicit-Ought ResearchRabbit Zotero Mendeley Mindgrasp Coconote Knowt ChatGPT-Claude; not-in-Dev-Audio-Video-11card-set edu-research-productivity-sublane.

StudyFetch is the "course materials in, full study system out" tool for students. Upload anything (PDF, PPT, DOCX, MP3, MP4, YouTube link, camera photo of whiteboard), and it ingests, transcribes, and chunks the content, then emits structured notes (summary/in-depth/comprehensive), active-recall flashcards with spaced repetition, practice quizzes, a 6–45 min podcast-style Audio Recap, and a Socratic voice/chat tutor (Spark.E) that answers only from your sources. Live Lecture Assistant​ records a class in real time, transcribes single-speaker audio, and links notes to your study set before class ends. Arcade​ turns the same material into game challenges; Study Calendar​ breaks deadlines into daily milestones; Essay Grader​ gives rubric-style feedback; Wolfram Alpha is wired in for STEM steps. It sits between NotebookLM (source-grounded chat, no flashcards/spaced-rep) and Quizlet (flashcards, no ingestion/tutor) and Mindgrasp (notes-only) — widest ingestion-to-tutor surface in the student lane, weakest on hard-STEM accuracy and free-tier generosity.

Symbolab takes a typed, pasted, or photo-scanned math problem and returns the final answer plus a numbered, rule-labelled worked solution — "factor here (difference of squares)", "apply chain rule", "row-reduce: R₂ − 3R₁". Coverage spans arithmetic, algebra, trig, pre-calc, single/multivariable calculus, linear algebra (matrices/vectors), differential equations, limits, series, and intro statistics, backed by 500+ topic calculators and a built-in graphing tool. The LLM layer writes the plain-English narration and powers the AI Chat follow-up + word-problem interpreter; the underlying transforms are symbolic, not guessed. It sits between Photomath (camera-first K-12) and Wolfram|Alpha (computation-first research): weaker than Wolfram on proof-level/deep symbolic work, stronger than Photomath on college calculus and desktop typed input.

Teachmate — UK-Built AI Teaching Co-Pilot: 160+ Curriculum-Aligned Tools for Lesson Plans, Worksheets, Slides & Reports (GDPR / DfE / CyberEssentials)

Wolfram|Alpha computes answers rather than retrieving prose. You type natural language ("integrate x^2 dx", "population of France / GDP per capita", "Mars opposition 2026"), it disambiguates via its NLU system, executes against the Wolfram Language engine + curated Wolfram Knowledgebase (20+ domains, real-time feeds for weather/stocks/astronomy), and returns pods of exact results, derivations, plots, and source-attributed data. In the LLM era its role is the deterministic compute+fact layer behind probabilistic models: the model formulates a query, Wolfram executes it exactly, and the result is woven back into fluent text. Stephen Wolfram called the combined system "the first truly large-scale statistical + symbolic AI system" — the neural net handles language, Wolfram handles anything that must be exact, computed, looked up, or visualised.