AP exams shape college admissions. But the best prep tools cost hundreds of dollars — or require hiring a tutor. StudentNest exists to change that: world-class AP, SAT & ACT prep with the Sage tutor, free for every student.
For high schoolers, AP prep hasn't changed in 20 years — thick review books, expensive tutors, or generic flashcard apps.
We built StudentNest to solve both problems. A platform that does what a great tutor does: explain why, not just what — then test whether it clicked.
Sage, our Sage Live Tutor, is the result of that idea: a study partner that's available 24/7, never gets impatient, and adapts to exactly where you're struggling — whether you're aiming for a 5 on AP World, a higher SAT score, or a better ACT composite.
Our Mission
“Make world-class AP, SAT & ACT prep accessible to every student — not just those who can afford a tutor.”
Active recall. Research consistently shows that actively retrieving information — not passively rereading notes — is the most effective way to move knowledge into long-term memory. Every StudentNest practice question and every Sage comprehension check is an act of retrieval.
Spaced repetition. When you answer a question wrong, StudentNest schedules it to reappear at increasing intervals. This technique has been shown in cognitive science studies to dramatically improve retention over weeks and months compared to massed practice.
Mastery-based progression. Rather than assigning the same material to every student, Sage tracks your mastery by unit and generates questions targeting your weakest areas first. This approach — rooted in Bloom's mastery learning framework — ensures that practice time goes where it has the most impact.
Comprehension verification. After Sage explains a concept, it offers an optional 3-question check to verify you understood the explanation — not just read it. This closes the loop between passive learning and active demonstration of understanding.
Every practice question goes through a rigorous 9-step quality pipeline
Topic Distribution
Questions match College Board’s official exam topic weights
Bloom’s Taxonomy
50%+ questions test application and analysis, not just recall
Scenario-Based
80%+ of questions present real-world scenarios, matching official exam formats
Misconception-First Distractors
Wrong answers target actual student misconceptions
Generation Stage
Gemini 2.5 Pro generates questions calibrated to exam level (cross-model validation in next stage)
Cross-Family Validation
A second model from a different family independently validates every question — defeats single-model blind spots
8-Criterion Review
Factual accuracy, single correct answer, distractor quality, cognitive level, exam alignment, scenario check, distractor distinctness, explanation quality
Difficulty Calibration
Empirical tracking ensures questions match real exam difficulty
OpenStax Grounding
Factual accuracy verified against free, peer-reviewed college textbooks
120+ original calibration question stems validated against College Board exam rubrics
StudentNest is not affiliated with or endorsed by the College Board.
Instant Feedback on Every Answer
Every MCQ and FRQ gets an immediate explanation of why the correct answer is right and why each wrong answer is a common misconception. No waiting, no answer keys.
Mastery Tracking Across All Units
Your dashboard shows per-unit mastery percentages, accuracy trends over time, and a visual heatmap — so you always know exactly where to focus your next study session.
Adaptive Study Plans
Each week, Sage generates a study plan that targets your weakest units first and adjusts recommendations as your scores improve. Set your exam date and the plan counts down with you.
Here's what the experience looks like — from your personalized study plan to practice questions with instant AI feedback.
Master Units 4 & 5 before Friday
Focus Areas
Unit 4: Revolutions
41% mastery — 12 MCQs recommended
Unit 5: Cold War
63% mastery — 8 MCQs recommended
Unit 2: Exploration
74% mastery — 5 MCQs to reinforce
Mon: Unit 4 MCQs
Tue: Unit 5 Review
Wed: Mixed Practice
Thu: Weak Topics
Prioritized weekly plan by unit
MCQ · Unit 5: Revolutions
Which of the following BEST explains why the French Revolution spread ideas of democracy across Europe?
Napoleon's military campaigns (1799–1815) carried Revolutionary ideals of liberty and nationalism into conquered territories, abolishing feudal privileges across Europe.
Instant Sage explanations
78%
Accuracy
342
Questions
12d
Streak
2,450
XP
Mastery by Unit
Accuracy Over Time
Per-unit mastery tracking
Outcome-Obsessed
We measure success in score improvements, not time-on-site.
Accessible First
Core features are free. Always. Premium is for students who want more — not a paywall for the basics.
Sage as a Teacher
Sage doesn’t just answer questions. It explains, checks your understanding, and adapts to your weak points.
AP Courses
SAT Prep
ACT Prep
16
Courses covered
24/7
Sage Live Tutor availability
Free
To start, forever
$9.99
Premium per module/mo ($79.99/yr)
A landing-page rebuild driven by real student feedback (“feels like another AI study tool”) plus a complete restructure of how courses are gated. Hero now demos a real wrong-answer flow on its own. CLEP/DSST handed off to PrepLion. Course visibility is now a per-bank quality gate (≥200 vetted questions before users see it). And a permanent E2E guard blocks the React-error class that surfaced the “hiccup” fallback page from ever shipping again.
Hero auto-shows the failure flow
Hero replaced static analytics screenshot with a live interactive question. Visitor doesn't even click — at 2.5s, the demo auto-clicks a wrong answer so they see the tension banner (“You're confusing media history with political diffusion”). Pause at first user click. Designed to flip the demo from “static product showcase” to “live mistake experience” in 5 seconds.
Quality gate: only courses with ≥200 vetted Qs are visible
Earlier we shipped questions with arithmetic errors and answer-key bugs. Students were marked wrong on questions where they reasoned correctly. Now, courses live behind a SiteSetting allowlist. A course shows up only after passing 7 deterministic gates (math, structural, distractor-leak, stimulus, source-attribution, figure, course-contracts) plus two cross-family LLM judges (Gemini + GPT-4o). 13 courses live, 16+ rebuilding to threshold autonomously.
CLEP & DSST → PrepLion
StudentNest serves AP/SAT/ACT only. CLEP and DSST seekers were trying to use marketing copy that didn't reflect us. Permanent 308 redirects on /clep-prep, /dsst-prep → preplion.ai. Footer carries a small “CLEP & DSST? → PrepLion” handoff link for honest cross-product navigation.
Permanent runtime-error guard
Beta 11.0 hotfix: a React hook-order bug in SageChat crashed the landing page on scroll. Existing tests asserted text visibility but never simulated user scroll. New permanent E2E spec walks the page like a user — load, scroll, click — and hard-fails on any uncaught pageError, hook-order error signature, error-boundary fallback, or same-origin 500. Runs on every staging gate. The class of bug that shipped Beta 11.0's crash is now blocked.
Sidebar: AP + SAT + ACT all visible together
Test user reported only AP courses showing in their sidebar, even though SAT/ACT were also vetted. Root cause: legacy track-lock (one signup track = one course family). Fixed: students prep for multiple exams; sidebar now shows all 3 group tabs regardless of signup track. Track only determines the default-selected tab, not a hard lock. FRQ Practice nav gated by current course family, not signup track.
Day 1 → Day 7 timeline + multi-Q tension section
Replaced 8 generic feature cards with a concrete time-to-readiness flow: Day 1 take diagnostic, Day 2 weak units surface, Day 3 practice only what matters, Day 5 timed simulation, Day 7 ready. Plus a “Most students think they're ready / Until this happens” section showing 5 questions, 3 wrong, confidence drops to 40%. Slightly uncomfortable on purpose — catches overconfidence early.
Five P0 quality fixes from real-student feedback during AP season — answer-letter distribution rebalanced (no more guessing “A”), explanations cut to 40-80 words with optional “Show full,” the practice-flow Quick Check disabled by default, the official AP World History FRQ library ballooned from 0 to 104 questions (plus 519 more across 13 other AP courses), and the session-feedback popup now actually fires (was dismissable at session #1; now triggers at sessions 1, 5, 10, 25, 50, 100, 200 and blocks dismiss until rated).
Answer Letter Re-Shuffle (All 34 Courses)
Audit found 86% of AP World History MCQ correct answers were “A.” Real CB exams use ~25% per option. Fisher-Yates shuffled every existing MCQ (and added a generation-prompt rule to prevent drift) so no course is now >30% any single letter.
523 Official AP FRQs (was 90)
Built a CB index-page scraper that hits each AP course's past-exam-questions URL and pulls the actual current PDF set. Extracted via Groq into structured JSON with rubrics + scoring guides. AP World History alone went from 0 to 104 official FRQs covering 2019-2025.
Concise Explanations + Show-Full Toggle
Old explanations averaged 800-1600 chars across courses (vs CB's typical 200-400). Tightened generation prompt to 40-80 words and added smart truncation — first ~280 chars show by default, “Show full explanation” toggle for the rare long ones.
Quick Check Disabled by Default
The post-wrong-answer 3-MCQ comprehension quiz was firing during AP-season cram and slowing students down. Now off by default; admin flag still available to re-enable.
Feedback Popup Actually Fires
Old popup showed once per (course + source + context) and was dismissable via Escape — only 5-8 unique users gave feedback in 5 weeks despite 40 signups/week. New: triggers at session 1, 5, 10, 25, 50, 100, 200 milestones, blocks dismiss until rated. Expect 5-10x more feedback volume.
Index-Page Scraper (CB-Compatible)
CB silently changed their PDF URL pattern in 2023 (added -set-N suffixes). Old hard-coded URLs returned 404. New scraper hits the past-exam-questions index page for each course, parses out current PDF URLs, downloads everything publicly available. 999 PDFs / 1.2 GB pulled across 24 AP courses.
Your projected AP score now tells you what to do next. Three ranked weak-unit recommendations appear right under your 1–5 prediction — each with a one-tap path to focused practice and a rough estimate of how many questions to bump into the next score tier. No more guessing what to study.
Score-to-Action Recommendations
Below your projected 1–5 score, the dashboard now shows your top 3 weakest units, each labeled with current mastery and a rough Qs-to-next-tier estimate. One tap lands you in focused practice for that exact unit.
Untouched Units Surface Too
Units you haven't practiced yet appear as “Untouched · start here” cards alongside weak-mastery units. We rank by boost potential — not just lowest mastery — so you build a complete prep, not a lopsided one.
Calibrated Question Estimates
“~12 Qs to next tier” isn't a guess — it's based on a conservative ~10 questions per 8 percentage points of mastery gain, calibrated against historical data. Floor at 5, cap at 50.
Same Score, Smarter Pathway
The 1–5 scaled prediction itself didn't change. What changed is the “OK, what now?” gap between “you're a 3.7” and actually doing the thing. Three concrete actions appear automatically — no extra clicks, no extra page.
Screen-reader fixes for every progress bar across the app, plus the foundation of the College Board FRQ ingestion pipeline (47 official PDFs downloaded across 8 AP courses). Plus a P1 batch from the bug-hunt sprint: cache self-heal, email compliance, mobile touch targets, flashcard explanation cleanup, and an escape-hatch from the “No FRQs available” dead-end for Premium users.
Progress Bars Now Screen-Reader Friendly
Every <Progress> across analytics, practice, diagnostic, flashcards, am-i-ready, and the daily-goal card now carries a contextual aria-label. Caught by axe in the post-deploy a11y scan (rule: aria-progressbar-name, severity: serious).
FRQ Library Foundation
Built the College Board FRQ ingestion pipeline: 47 official past-exam PDFs across 8 AP courses (Bio, Chem, Calc AB/BC, Stats, Physics 1, US History, World History) downloaded with rate-limiting and idempotent caching. Extraction stage uses Gemini 1.5 Pro PDF-native input. Currently waiting on a refreshed API key; once unblocked, ~200 official FRQs land in the library.
FRQ Empty-State Escape Hatch
Premium users picking an AP course with no seeded FRQs no longer dead-end with “No FRQs available.” Now offers a “Generate FRQ with Sage” CTA into the Sage-FRQ flow already wired in /practice — same rubric scoring, same difficulty calibration.
Flashcard Explanation Sanitizer
Flashcards no longer leak the source MCQ's “Why A is correct / Why B is wrong” scaffolding. The helper existed; just wasn't connected to the render path. Now it is.
Cache Self-Heal
Sage response cache no longer crashes on a single corrupt entry. JSON.parse failures self-heal — bad entry deleted, request falls through to fresh Sage call. Previous behavior: every request hitting the same cacheKey would 500.
Email Compliance + Deliverability
Every commercial email now includes List-Unsubscribe (RFC 8058 one-click) + reply-to. CAN-SPAM 5(a)(5) compliant + improved Gmail/Outlook inbox placement. Verification + password-reset opt out via { transactional: true }.
Three reliability fixes you won't notice until they would have bitten you — Stripe webhook no longer 500s on transient errors (silent revenue loss vector closed), failed card renewals now properly downgrade after Stripe's smart-retry exhausts, and Sage chat's streaming endpoint can't hang anymore. Plus two honest-copy fixes: landing page now matches /pricing on free-tier limits, and AP-prep's FRQ feature now correctly badges as Premium.
Webhook Reliability
Stripe webhook now returns 200 with structured error info on transient failures instead of 500. Reason: a 500 triggers Stripe's 3-day retry loop, but if the same transient error recurs, Stripe eventually marks the event permanently failed — meaning a paid user never gets their tier flipped. The hourly stripe-reconcile cron is the safety net.
Failed Renewal Handling
When Stripe fires `invoice.payment_failed` after exhausting smart-retry (4th failed attempt), the matching ModuleSubscription now flips to `past_due` so gating can prompt for a card update. No more “paid for Premium 60 days ago, card died, still has access” situations.
Sage Chat Can't Freeze
Sage streaming endpoint now caps the initial Groq response at 30 seconds via AbortSignal.timeout. Previously, a hung Groq endpoint would leave Sage chat's typing indicator running forever — fix avoids the worst-case UX.
Honest Free-Tier Copy
Landing-page free-tier card no longer says “Unlimited MCQ practice” (which directly contradicted /pricing's “3 sessions/day” cap). Now reads “3 practice sessions per day” — same number across landing + pricing + product. Cuts the “I felt deceived” support thread.
FRQ Premium Badge
AP-prep page now badges the “FRQs with rubric scoring” feature as Premium and clarifies free-tier gets one full FRQ to try. Reduces refund risk from users assuming unlimited FRQ on free plan.
Iteration 1 P0 Closeout
Comprehensive bug-hunt sprint Phase 1 (10 parallel agent scans) found 660 candidates; after deduping + verifying false positives, 6 real P0s identified. All 6 now shipped — Beta 7.2 caught 3 cold-start bugs, Beta 7.3 caught the remaining 3. Next iteration targets the P1 backlog (~17 items).
Reliability hardening across the dashboard, billing, and study-plan endpoints — fewer 5xx errors during fresh-isolate cold-starts. Plus the cross-module “you unlocked SAT/ACT too” nudge and edge-persistent rate limiting that actually works against anonymous flood traffic.
Cold-Start 500 Hardening
Three more endpoints (study-plan, mastery-tier-ups, billing-status) now wrap each Prisma call in a per-query safe-fallback wrapper. Same pattern that fixed the dashboard cold-start in Beta 7.0 — a single fresh-isolate hiccup no longer takes down the whole page.
Cross-Module Discovery Nudge
Premium users (any module) now see a post-session card suggesting they try a different exam they unlocked — 3 quick questions, 2 minutes. Drives discovery of the all-access entitlement that Beta 7.1 made real.
Anonymous Flood Protection
Migrated 3 IP-keyed rate-limit gates (check-verified, am-i-ready, anon Sage) from in-process Map (broken on Cloudflare Workers) to native Cloudflare Rate Limiting bindings. Persists across isolate recycles — actually rate-limits anonymous traffic now.
Stuck-Onboarding Recovery (24h + 72h)
Two-touch email sequence catches users who signed up but never finished setup. Email 1 at 24h is friendly (‘we saved your spot’); Email 2 at 72h is AP-season urgency-aware. Hard cap of 1+1 emails per user, lifetime.
Pricing Copy Cleanup
All remaining “per module” references replaced with all-access language across FAQ, JSON-LD schema, and Sage’s system prompt. CLEP dropped from StudentNest user-facing copy (lives on preplion.ai).
Early-Stage Predicted Score
/analytics now shows a tentative predicted-score panel for users with 1+ answer (instead of staying blank until first mock exam). The first-answer reward modal’s ‘see my predicted score’ CTA actually delivers something visible from day one.
One Premium subscription now unlocks every exam. Plus a full landing-page conversion pass — sharper headlines, exam-specific demo content, honest claims, and a tighter pricing page.
Premium Is Now All-Access
One $9.99/mo subscription unlocks every exam on StudentNest — AP, SAT, ACT — including practice on every course. Sign up via whichever exam you start with; you get the whole platform. (Existing subscribers automatically gained access to everything else, no action needed.)
Landing Pages Rewritten for Clarity
ACT, AP, SAT, and pricing pages all got a conversion-audit pass: sharper outcome-driven headlines, exam-specific demo content (no more AP World units showing on the SAT page), pain statements, and a tighter pricing comparison table.
Honest Score-Gain Language
Replaced absolute score-jump claims (e.g., ‘1050→1350’) with typical-pattern framing (‘1050 → 1150 → 1250+’). More credible for parents, safer legally, more inclusive of students starting at any level.
Pricing Page De-cluttered
Removed the “subscribe to multiple modules independently” framing that was creating cognitive load. Free tier no longer says both “unlimited practice” AND “3 sessions/day”. Refund + cancel reassurances consolidated to one trust line.
FRQ Differentiator Surfaced
AP page now leads its second feature with “Practice FRQs with rubric scoring on real AP standards” — the actual differentiator vs Khan/Quizlet/ChatGPT. Was previously buried behind generic MCQ copy.
Pain Statements Added
AP and SAT pages now lead with the user-pain insight: students plateau because they re-read what they already know, not because they don't study enough. Sage drills only the points you're actually losing.
AP exam season is here — this release is built around getting you from sign-up to your first practice answer faster, and giving you a moment of celebration when you do. Plus reliability fixes for the dashboard, accessibility upgrades across the marketing site, and an honest rewrite of the ACT page.
First-Answer Celebration
Submit your very first practice question and we'll quietly celebrate — with a one-time modal that tells you exactly how much progress your predicted score will make as you keep going. Designed to land at the highest-engagement moment in the entire product.
AP Season Countdown
Live calendar-aware urgency banner across every marketing page — counts down the days until the AP exam window so you know exactly how much time is left. Auto-dismissable, never repeats once closed.
Honest ACT Page
Rewrote the ACT prep page hero to remove unsupported claims (no more ‘3-5 point lift’ without data). Sage Coach voice practice now appears as a fourth pillar on the landing page — previously it was completely undiscoverable.
Dashboard Reliability
Fixed cold-start 500 errors that occasionally made the dashboard predicted-score card vanish for the first 2-5 seconds after sign-in. Per-query fallback wrappers mean a single slow Prisma query no longer takes down the whole panel.
Flashcards Get Rich Rendering
Markdown tables and LaTeX formulas now render properly on flashcard front, back, and explanation surfaces. Catches up flashcards to where practice and mock-exam already were.
Accessibility Across Marketing
Logo ‘Prep’ suffix, AP/CLEP toggle chips, and pricing refund line all now meet WCAG AA contrast (4.5:1+) in both light and dark mode. Screen-reader users and low-contrast viewers get a measurably more readable experience.
Reliability + accessibility pass — a flurry of small fixes you'll feel without noticing. Better error pages, clearer login feedback, smoother first paint, and security headers that score A on every external scanner.
Branded Error & 404 Pages
Stale links and unexpected errors now land on a friendly StudentNest page with three escape hatches (Dashboard / Home / Pricing) and a quotable reference code for support — no more dead-end Next.js fallbacks.
Faster-Feeling Dashboard
The predicted-score card now shows a skeleton in the right shape instead of a bare spinner, and reassures you when the worker takes longer than usual. No more layout shift when content arrives.
Login & Sign-Up Recovery
Wrong password? You now get a clear, persistent banner with a one-click reset link. Email already registered? Direct link to log in instead. Toasts disappear; banners stay until you act.
Sage Coach Voice Restored
A regression in our security headers was blocking microphone permission for Sage Coach voice practice. Fixed and verified.
Onboarding → Premium Path
Picking Premium during onboarding now reliably routes to /billing instead of bouncing to /dashboard — a race condition in the JWT refresh sequence is fully closed.
Accessibility Polish
Tighter color contrast across muted text (WCAG AA-compliant), proper heading order on prep pages, password show/hide gets a screen-reader label, and inline links carry an underline at rest — not just on hover.
Question rendering + sign-up polish — markdown tables and LaTeX now render correctly in practice, mock exam, and diagnostic. Sign-up copy reflects your chosen prep track.
Rich Question Rendering
Physics, chemistry, and statistics questions with data tables now display as proper tables, not raw pipes. Math formulas render with LaTeX. Applied to practice, mock exam, diagnostic, and FRQ.
Track-Aware Sign-Up
Starting from the SAT, ACT, or CLEP landing pages now surfaces SAT / ACT / CLEP-specific framing on the sign-up screen instead of defaulting to 'AP exam journey.'
Wall of Fame Public View
The public leaderboard preview on /wall-of-fame no longer throws 401s in the browser console — anonymous visitors see the real top-10. Personal rank + XP stay gated behind auth.
Cleaner Pass-Rate Links
The /pass-rates reference table no longer links to difficulty-guide pages we haven't built yet. Data rows stay for honesty; links appear only when the destination exists.
Conversion funnel overhaul, 5 new AP courses in preparation, and question-quality guardrails responding to community feedback.
Diagnostic → Focused Practice Bridge
After your diagnostic, see your predicted AP score + a 2-minute focused session on your weakest unit in one click. Insight lands on action, not a dead end.
Question Quality Guardrail
New anti-ambiguity rule rejects exam questions where multiple answers could be defensible — the 'primary / main / most important' trap that Reddit users flagged. Applied to both the generator prompt and the second-pass validator.
5 New AP Courses (In Preparation)
AP Human Geography, AP U.S. Government, AP Environmental Science, AP Precalculus, and AP English Language scaffolded end-to-end. AP HuGeo and AP Environmental Science already at 500+ College Board-grounded questions; the rest visible to admins while Phase C completes.
Mobile Bottom Nav + Haptics
Four-tab bar for Home / Practice / Mock / Progress on mobile, hidden during exams. Haptic feedback on every answer — short success pattern on correct, error pattern on wrong. Works on Android; graceful no-op on desktop + iOS pre-18.
Trial Days-Remaining Banner
A 3-day countdown banner appears at the top of the app once your free trial is within its final 72 hours. Severity escalates on the last day with a direct path to upgrade, so you never miss the renewal window by accident.
Funnel Instrumentation + Paywall Repair
Fixed a silent race condition that was blocking 100% of our coach-funnel metrics. Locked-insight overlay now surfaces your full predicted breakdown only after you start a trial, turning the diagnostic into the cleanest path into the paid tier.
16 courses at 500+ CB-validated questions. Sage Coach oral training. Parent Invite + retention engine. Auto-quarantine on student reports.
Per-module subscriptions, SEO overhaul, content optimization, and strict course filtering.
Per-Module Stripe Subscriptions
Independent $9.99/mo subscriptions for AP, SAT & ACT. Subscribe to exactly the module you need. Cancel or reactivate each independently.
Module-Locked Sidebar
Sign up from /ap-prep → see only AP courses. SAT → only SAT. No more seeing all 16 courses — focused experience from day one.
SEO & Content Overhaul
JSON-LD structured data, OG images, XML sitemap, 4 dedicated prep landing pages with study flows, outcome messaging, and parent trust sections.
Outcome-Driven Copy
Every page now leads with results: score improvements, prep timelines, cost savings. Comparison tables vs tutors and ChatGPT on pricing page.
Previous releases (Beta 1.0 – 2.0) introduced the core Sage Live Tutor Sage, daily streaks, spaced repetition, voice input, automated question bank seeding, admin dashboard redesign, annual billing, and a 7-day money-back guarantee.
Have a question, found a bug, or want to suggest a course? We read everything — reply within 24 hours.
[email protected]For pricing details, see the Pricing page.
AP® is a trademark registered by the College Board, which is not affiliated with, and does not endorse, this product or site. SAT® is a trademark registered by the College Board, which is not affiliated with, and does not endorse, this product or site. ACT® is a registered trademark of ACT, Inc., which is not affiliated with, and does not endorse, this product or site.
StudentNest is an independent educational technology platform. All practice questions and study materials are original works created by StudentNest and AI models — not reproduced from any official exam publisher.