1 Value Proposition Canvas
Three buyers, three different jobs. Pick an audience — the left side is what they live with, the right side is what Amitalux does about it.
Jobs
- Plan and differentiate lessons; set and mark homework/tests; give useful feedback fast
- Keep every student thinking rather than copying; know who is stuck and where
- Communicate with families without writing the same message 30 times
- Stay inside school AI policy — and be seen as a competent, fair professional who keeps authority over grades and judgement
Pains
- AI does assessed work: 55% of K–12 teachers in the Jun 2026 NPR/Ipsos poll called AI “mostly a shortcut”; 54% said it harms critical thinking; 39% now require more handwritten/in-class work
- No guidance: 82% had no formal AI guidance (Gallup/Walton, Feb–Mar 2026, n=2,069); ~half of schools have no AI policy
- Marking load — especially maths, where the error is buried mid-solution
- Checking AI output eats the time it saves (56% of UK teachers, Up Learn 2026); fear of entering student data into an unsecured tool
Products & services
- Per-class Class AI for research, homework and test correction
- Teacher enable / pause per class; teacher-defined approved sources and help levels
- On-device maths worksheet review that circles the first wrong step
- Teacher review before any family reply is shared
Pain relievers & gain creators
- The assessed-work boundary is the product, not a setting — and the teacher can pause Class AI entirely
- Maths review points at the first wrong step instead of re-doing the worksheet — the teacher keeps the judgement; photo/steps stay on device
- Class-aware help: explains, asks a guiding question, gives a worked example at the teacher-allowed level
- One connected workspace (dashboard, classes, assignments, calendar, study tools, discussion) instead of another chat tab; clear history for actions, reviews and changes
Fit — where Amitalux beats the set
MagicSchool / Brisk / Diffit / Eduaide are mainly teacher-productivity tools; SchoolAI / Flint / Khanmigo are student-chat tools where the teacher monitors after the fact. Amitalux is the only one in this set whose core promise is teacher-governed per class, in the workflow, with assessed work excluded by design — a design claim from the mission/website, not yet proven with real students.
Jobs
- Adopt AI without a board-level privacy incident; roster reliably (OneRoster / SIS)
- Prove usage and value to the board; keep costs predictable per student
- Meet FERPA/COPPA and state privacy law; give principals one consistent policy instead of 40 teacher tools
- Be the leader who brought AI in safely — and defend the purchase in a public meeting
Pains
- Privacy/compliance exposure is superintendents’ top fear
- Free-now-pay-later tools; quote-only pricing with no budget number without a sales cycle
- Student chat the district can only oversee by reading every transcript — a safeguarding and workload problem
- Procurement takes 6–18 months; board approval can slip a purchase a full school year. No vendor in this set publishes a peer-reviewed RCT proving learning gains
Products & services
- OneRoster 1.1 roster import
- District dashboard that is aggregate-only — no student names, question text, or gradebook records
- Teacher-governed Class AI with school-staff review for safety concerns and consequential decisions
- Optional Library Books add-on
Pain relievers & gain creators
- Aggregate-only dashboard removes the “district reads children’s chats” objection and the transcript-review workload at a stroke
- Published per-student pricing removes the quote-only budget blocker; a free, time-boxed pilot removes “pay before proof”
- One platform for classes, assignments, study tools and family communication — one system instead of a teacher tool + a tutor + a parent app
- Safety review before ordinary subject help (per amitalux.ai/schools); auditable history without reading transcripts
Fit — and an honest gap
On governance, Amitalux is narrower and more defensible than SchoolAI/Flint (full chat visibility) and more student-facing than MagicSchool/Brisk. On price transparency it beats every quote-only incumbent. On evidence it is currently behind — Khan cites Tier 1/2 studies, SchoolAI claims ESSA Level III — Amitalux has no real-student evidence yet and must earn it in the pilot.
Jobs
- Know how my child is doing and what they should do next
- Help with homework without doing it for them
- Know what AI my child is using at school and what it can / cannot do; reach the teacher through a proper channel
- Feel my child is safe — not watched, profiled or exposed — and feel respected as a partner
Pains
- No visibility: homework happens in a chat the parent cannot see, or AI is only mentioned when something goes wrong
- Privacy fear: who reads my child’s messages, what is stored, is it used to train models?
- Cheating worry: UK polling — parents reporting a child using AI for homework rose 47% → 57% in ten months (Bett/YouGov via Public First)
- Being asked to pay: Khanmigo charges parents $4/mo or $44/yr. Getting a raw AI reply about their child with no human check
Products & services
- Family portal showing activity counts and shared progress reports
- Guardians get no chat transcript or files
- A teacher must review before sharing a reply
- Access limited to the appropriate child and approved information
Pain relievers & gain creators
- No transcript / no files is a privacy feature, not a missing feature — the opposite of SchoolAI Scale / Flint full-chat visibility
- Teacher review prevents an unreviewed AI message reaching a family
- Counts + progress reports give a parent something concrete to talk about at home without policing homework
- Free to families (proposed) removes the Khanmigo $4/mo parent-pay friction
Fit — clearest uncontested space, with a caveat
No competitor in this set offers a family portal with this exact boundary (counts + reviewed reports, no transcript, free). Caveat: the portal is listed on amitalux.ai as “planned” / still to be built, so it cannot be sold as live until it is verified.
Differentiators at a glance
| Differentiator | Who else does it? | Edge / caveat |
|---|---|---|
| Teacher enable/pause per class; teacher-defined sources & help levels | MagicSchool Enterprise tool controls; SchoolAI persistent guardrails; Brisk curriculum grounding — all admin/district-level | Control sits with the classroom teacher, per class. Caveat: unproven at scale with real students. |
| No assessed work — Class AI helps with research/homework/test correction without doing it | Flint’s Sparky and Khanmigo guide rather than give answers (tutoring style) | Stated as a product rule for assessed work. Needs a published, testable definition and pilot evidence it holds. |
| On-device maths review circles the first wrong step | No competitor in this set claims on-device maths marking or first-wrong-step circling | Strongest privacy + pedagogy claim. Caveat: devices/maths scope and what leaves the device not yet published. |
| Family portal: counts + reviewed reports; no transcript, no files | None in this set. Khanmigo parent route is paid ($4/mo) | Uncontested. Caveat: portal is “planned” — not yet built/verified. |
| District dashboard aggregate-only | Opposite: SchoolAI Scale (full chat observability), Flint (admins read every conversation) | Only “no surveillance by design” district view in the set. Trade-off: districts wanting student-level intervention data will see a gap — sell it as a deliberate boundary, with detail at teacher level. |
| OneRoster 1.1 import · Grades 6–12 focus · Library Books add-on | OneRoster/Clever parity elsewhere; all competitors are K–12 or broader; no direct library add-on found | Parity on rostering (essential), depth in 6–12, and a clean optional revenue line that never gates the core platform. |
2 Proposed pricing model
All prices in this section are proposals grounded in the competitor anchors — they are not Amitalux prices (Amitalux publishes no price) and not competitor prices. Do not present them to a buyer as market rates.
0 — Pilot · 2026–27
One term, up to 16 weeks · one school · up to 10 teachers / 300 students
- Everything in School tier + signed DPA & parental notice template
- OneRoster 1.1 import (or CSV if OneRoster isn’t ready — state which at signing)
- Family portal where built — if not live, the pilot must say so and exclude it
- Onboarding + 2 live PD sessions; named pilot contact; exit pack with deletion confirmation
- Library Books: free limited catalogue
Conditions: evaluation, not a free year. Success criteria agreed in writing before Day 1 (e.g. % teachers using Class AI weekly, assessed-work boundary incidents = 0, family-portal use, teacher time on maths review). No student data used to train models.
1 — Teacher
or $9/mo billed monthly ($108/yr). Annual saves $29 (27%).
- Class AI for that teacher’s classes (enable/pause, approved sources, help levels)
- On-device maths worksheet review; class dashboard
- Family portal for that teacher’s classes (review-before-share)
- Email support; self-paced PD
- No OneRoster, no district dashboard, no SSO
Family
Parents/guardians of an enrolled student
- Own child only: activity counts, shared progress reports, teacher-reviewed replies
- No transcript, no files, no charge, no ads, no upsell
Khanmigo charges parents $4/mo — charging guardians would contradict the privacy/trust positioning.
2 — School
Minimum $2,400/yr — the school pays $6 × enrolment or $2,400, whichever is higher. Annual only, July–June.
- Everything in Teacher for all teachers; teachers/admins not charged separately
- OneRoster 1.1 import; SSO (Google/Microsoft); standard DPA; admin controls
- Family portal school-wide; school aggregate dashboard (no names/question text/gradebook records)
- Onboarding + 2 live PD sessions/yr; email + chat support
3 — District
$5 for 1,001–5,000 students · $4 for 5,001–15,000 · $3.50 for >15,000. Minimum $6,000/yr. Students = enrolled 6–12 students covered.
- Everything in School for all schools, plus district aggregate-only dashboard
- Custom DPA (redlines included); SIS/LMS integration support
- Dedicated customer success manager; white-glove onboarding; priority support
- Unlimited live PD (within reason, stated in contract); annual usage/evidence review
Library Books add-on
25% of the $6 School rate. Minimum $300/yr per school, capped at $1,500/school/yr (a 1,000-student school pays $1,500). District rate: $1.25/student/yr, capped at $1,250/school/yr when bought district-wide.
No competitor publishes a comparable add-on price, so this is priced as a proportion of the core platform, not by analogy. Sold only as an add-on to a paid School/District tier (or included in the Pilot) — never standalone, never required for Class AI. Free limited catalogue in the Pilot. If catalogue/licensing terms aren’t settled, sell “pilot catalogue only” and take no paid add-on orders yet.
Billing, discounts & what’s not included
- Annual only for School/District, billed July–June (mid-year start pro-rated to next 1 July — a January start pays ~50% for Jan–Jun)
- Multi-year: 2-year = 10% off; 3-year = 15% off, price locked for the term
- Founding-pilot converters: 25% off Year 1 if converting by 30 Jun 2027, with the Year 1 rate locked as the Year 2 rate (e.g. School at $4.50/student in Year 1, $6 in Year 2)
- Purchase order accepted for School/District; card for Teacher. No per-chat, per-generation or per-token charge on any paid tier — unlimited Class AI use within the licence, with a fair-use guardrail in the contract
- Not in the price: devices; SIS/LMS licences the district already pays for; bespoke curriculum builds (quoted separately); Library Books (add-on above)
- ESSA — stated honestly: Amitalux is not ESSA-validated yet (unlike SchoolAI’s Level III claim). Do not claim Title I–IV eligibility until an evaluation exists; the pilot is designed to generate that evidence
Why these numbers — seven design principles
- Per-student is the district language; per-teacher is the pilot language. Districts budget per student (Khan $10, SchoolAI ~$4.50 contract, Google $6, Ed.ai $10). Teachers buy per teacher (MagicSchool $99.96, Brisk ~$99.99 reported, Diffit $149.99, Eduaide $49.99). Price both — and make the per-student rate cheaper than buying teacher seats for everyone.
- Below the $10 Khan anchor, at the SchoolAI contract level, above the $0 floor. $6 / $5–$4 is defensible as “whole platform, not a tutor” without looking like a discount vendor.
- Teacher price below MagicSchool/Brisk, above Eduaide. $79 signals “more than a planning tool ($49.99), less than the big toolbox ($99.96)” — plus student-facing Class AI and a family portal Eduaide doesn’t have.
- Families never pay. See Family tier above.
- No hidden AI/token charges. Competitors cap generations or launches and charge for headroom. A pilot-stage vendor that meters AI use will be blamed for every “it stopped working” moment.
- Annual for institutions, monthly only for individuals. A monthly district price would not fit a purchase order.
- The pilot is free, time-boxed and evidence-generating — because Amitalux has no real-student evidence yet. Its job is to produce the usage, teacher-time, boundary-holds and family-satisfaction evidence that justifies the 2027–28 paid price.
Worked examples — try your own size
Proposal arithmetic only. Change the student count to see the Amitalux proposal next to the published anchors at the same size.
Reference points from the report: 300-student school → $2,400 (minimum binds; Khan $3,000, Flint $4,000 at that size) · 5,000-student district → $25,000 (Khan $50,000 at Starter rate; SchoolAI $22,500) · 10,000 → $40,000 · 20,000 → $70,000. Single teacher → $79 vs MagicSchool $99.96, Brisk ~$99.99 reported, Diffit $149.99, Eduaide $49.99.
Per-student price — where Amitalux sits
Published / contract anchors only, USD per student or user per year. Flint shown at its effective $13–$20/user range; SchoolAI is a single contract, not a list price.
3 Competitor pricing comparison
Prices as found 15 November 2026. QUOTE = not published, must be quoted — no figure invented. ESTIMATE / CONTRACT = secondary estimate or a single negotiated contract, not a list price. Re-check each official page before a purchase.
| Competitor | Price | What the buyer gets for that price | Privacy / data terms (as claimed) |
|---|---|---|---|
| MagicSchool magicschool.ai/pricing — verified live 15 Nov 2026 | Free $0 · Plus $12.99/mo or $99.96/yr ($8.33/mo annual) per teacher · Enterprise QUOTE | Free: 80+ teacher tools, generation-capped, no student tools. Plus: unlimited generations/history/editing, 50+ student tools, Student Rooms, insights. Enterprise: custom DPA, SIS/LMS (Clever, ClassLink, Canvas, Schoology), SSO, district alignment, tool controls, advanced moderation/dashboards, CSM, white-glove onboarding, custom PD. Also via purchasing cooperatives (UEN, ESC, Shop DOE NYC, MISBO, ESUCC, BuyBoard). | FERPA, COPPA, SOC 2, GDPR, Ed Law 2-D; Common Sense Privacy Certified. |
| Google Classroom / Gemini Education edu.google.com — verified live via fetch 15 Nov 2026 | Fundamentals $0 · Education Plus $6/user/yr · AI Pro add-on: price not shown / not verified | Fundamentals: Classroom, Docs/Drive/Meet, Gemini, Gemini Notebook, Gemini in Classroom, 100 TB pooled storage. Plus: Gemini in Docs/Sheets/Slides/Vids/Forms, advanced security/analytics, higher Notebook/Studio limits, extra storage. 25% off Plus on 50–999 licences, 1-yr commitment. Older “$3 Standard / $5 Plus” citations are out of date. | Enterprise-grade data protection; data not used to train AI models (per Google pages). |
| Khan Academy / Khanmigo support.khanacademy.org (updated 3 Nov 2026); khanacademy.org/schools/pricing | Core $0 · Teachers $0 · Learners/Parents $4/mo or $44/yr (US only) · Districts Starter $10/student/yr (≤1,000) · Enterprise QUOTE (>1,000) | Starter: district reporting, Khanmigo tutor (gr. 3–12), teacher-directed rostering, on-demand PD; US only, credit-card only, NCES ID required, terms non-negotiable, no PO path. Enterprise: automated Clever/ClassLink rostering, SSO, live PD, admin reporting. Note: older $35 / ~$15 district figures are superseded — $10 is the current official anchor. Nonprofit; price subsidised by philanthropy/Microsoft. | Microsoft Azure OpenAI infrastructure; district DPA via Enterprise terms (Starter terms non-negotiable). |
| SchoolAI schoolai.com/pricing — verified live via fetch 15 Nov 2026 (no prices published) | Free Trial $0 (5 Space launches/yr) · Pro QUOTE · Scale QUOTE · CONTRACT $4.50/student/yr — Linn-Mar CSD, 7,700 students = $34,650/yr, Jan–Jun 2026 | Free: Spaces, Dot chat, Mission Control, safety alerts. Pro: unlimited launches, SIS rostering, LMS integration, admin alerts/reports, live PD, CSM, 50 MB uploads. Scale: Student Portals & Sidekick, Sidekick in Canvas, student-level insights, custom standards, LLM audit traces, custom DPA redlines, full chat observability. Secondary estimates range $4–$10/student/yr — the spread itself shows the real price is negotiated and opaque. | FERPA, COPPA, SOC 2 Type II; ESSA Level III (Title I–IV eligible); 1EdTech / Common Good. |
| Brisk Teaching Third-party checks of briskteaching.com/plans, May–Nov 2026 — no live official price verified in this research | Free $0 · Educator Pro REPORTED ~$99.99/yr — flag: shown only inside the extension; $9.99/$10/$14.99/mo also circulate · Premium QUOTE, per student · Intelligence QUOTE, per student | Free: 20+ tools in Docs/Slides/Forms/Classroom/YouTube/Canvas, Boost (core). Pro: no caps, Turbo models, Inspect Writing / AI-detection + Replay. Premium: 35+ tools, admin dashboard, rostering, curriculum alignment, PD. Intelligence: outputs grounded in the district’s adopted curriculum, white-glove setup. | 93% (some sources: 95%) Common Sense Privacy Rating; FERPA/COPPA; SOC 2 Type II described as “working toward” — do not state it as held. |
| Diffit Checked 2 Oct 2026 via conversiongems.com and flintk12.com comparison | Basic $0 · Premium $14.99/mo or $149.99/yr · For Schools: flat annual by enrolment QUOTE | Basic: leveled texts, questions, vocab, PDF, 90-day history. Premium: Google/Microsoft/Canva export, standards/MTSS alignment, full history, graphic organisers, Diffit Chat. Schools: all staff, admin dashboard, free PD, PO payment. First-year teachers: free Premium through 30 Jun 2027; new accounts: 60-day trial. Teacher-only — students never log in; the $149.99 sets the ceiling for a teacher-only price. | “No student logins, no student AI … collects zero student data” (pricing page). |
| Curipod Conflicting sources — no verified official price. Toolentra review Nov 2026; SourceForge aggregator pages | Free $0 (session-capped) · Premium UNVERIFIED $7.50/mo annual reported by some aggregators; other Nov 2026 reviews say there is no published individual paid tier · School & District QUOTE | Free: polls, drawings, word clouds, AI lesson generation. Paid: unlimited sessions, 2,300-character responses, custom rubric feedback, progress reports, shared folders, SSO. The $3,999/school aggregator figure is unverified/legacy — do not use it. | FERPA, COPPA, GDPR claimed. |
| Flint flintk12.com comparisons — checked 30 Sep & 9 Nov 2026 | Free (teachers + students, max 3 teachers/admins per workspace) · School $3,000/yr (150 users), $4,000 (250), $6,500 (500); custom >500 | Sparky tutor — activities from the teacher’s own materials, asks questions rather than giving answers. Paid: unlimited teachers/admins, SIS/LMS sync, school-wide analytics, CSM, unlimited PD, school AI policy/settings. Closest published-price analogue for a school-wide student-AI platform. | Teachers/admins can read every conversation; self-harm alerts cannot be switched off. ESSA Level IV claimed. |
| Eduaide.Ai Verified 18 Oct 2026 on eduaide.ai/pricing via thesaninfo.com round-up | Free (15 generations/mo) · Pro $5.99/mo or $49.99/yr · Schools/Districts QUOTE | 100+ teacher tools grounded in named pedagogy (5E, UDL, Backward Design, DOK); Feedback Bot; document-chat. Teacher-facing only. | FERPA/COPPA for school/district tier. |
| Ed.ai ed.ai/en/pricing | Teacher Pro $49.90/mo (first month $24.90) · School $10/student/yr (volume rates from 1,500 students) | Grading/feedback-focused AI; every teacher included, school analytics dashboard, LMS integration, PO accepted, free 10-week pilot. | — |
| ChatGPT for Teachers / Microsoft Copilot / Claude for Teachers Indexed sources Aug–Nov 2026 — see Sources | ChatGPT for Teachers $0 for verified US K–12 educators — free-through date conflicts: June 2027 vs June 2028 · Copilot Chat $0 with A1/A3/A5; full Copilot $18/user/mo academic (flag: confirm with reseller) · Claude for Teachers free for a year for orgs signing up by 30 Jun 2027 | General-purpose drafting/planning workspaces. ChatGPT for Teachers is staff-only, “not for students.” These $0 assistants cap what any teacher-only tool can charge — but they are not governed student platforms. | Data not used for training by default (OpenAI); tenant-isolated, not used to train foundation models (Microsoft). |
| Amitalux Schools — PROPOSED | Pilot $0/term · Teacher $79/yr ($9/mo) · School $6/student/yr (min $2,400) · District $5 / $4 / $3.50 (min $6,000) · Family $0 · Library Books +$1.50/student/yr (min $300, cap $1,500/school) | Whole platform (classes, assignments, study tools, discussion) + Class AI (teacher enable/pause, no assessed work) + on-device maths first-wrong-step review + family portal (counts + reviewed reports, no transcript/files) + OneRoster 1.1 + aggregate-only district dashboard. | By design: aggregate-only district view; family portal no transcript/files; teacher review before sharing; privacy holds on unsafe uploads. FERPA/COPPA/SOC 2 to be verified before paid sale — not claimed yet. |
Not priced in depth: Canva for Education (free for eligible K–12) and Wayground/Quizizz (Basic + School-and-District only on its US page) — adjacent, not governed student-AI platforms. Do not quote prices for them.
4 Conclusion — why this model is best for Amitalux Schools
In one line: free pilot in 2026–27, then a transparent per-student price ($6 school / $5–$3.50 district) with a $79 teacher tier, families free, and Library Books as a capped add-on.
1 · It fits the stage
Charging $10/student or a quote-only Enterprise fee before portals, integrations and privacy controls are verified would put price ahead of proof — and invite the first DPA review to kill the deal. A $0, one-term, criteria-agreed pilot buys the only thing Amitalux lacks: real-student evidence. The 25% founding discount and Year-1 rate lock then convert pilot goodwill into paid revenue without a second sales cycle.
2 · It fits the positioning — and turns it into a price
The two most expensive things competitors sell are surveillance (SchoolAI Scale’s full chat observability, Flint’s read-every-conversation) and opacity (quote-only Enterprise everywhere). Amitalux sells the opposite — and a published price is the commercial form of that same transparency. A district can put $6/student in a January budget without a sales call; it cannot do that with MagicSchool, SchoolAI, Brisk, Diffit or Curipod. Family-free is the same logic.
3 · It fits the buying cycle
Districts plan Sep–Nov, budget Nov–Jan, approve Jan–Mar, procure Feb–May, and spend Jul–Aug on a July–June fiscal year; a purchase that misses the board cycle slips a full school year. A free pilot in the Sep 2026–Jan 2027 window (or the Mar–Apr 2027 evaluation window) produces evidence in time for Jan–Mar 2027 budget decisions and a Jul–Aug 2027 PO for 2027–28. The $79 Teacher tier runs in parallel as the bottom-up route — a teacher who pilots alone becomes the internal champion for the school purchase, the same conveyor belt MagicSchool’s $99.96 Plus was built for, at a lower price.
4 · It is grounded, not guessed
Every rate sits on a published anchor: $79 is 21% below MagicSchool/Brisk and 47% below Diffit; $6 is below Khan ($10) and equal to Google Education Plus ($6/user); $5/$4 is at/near SchoolAI’s real $4.50 contract and below Flint’s $13–$20/user; the $2,400 / $6,000 minimums protect onboarding/PD cost the way Flint’s $3,000 floor does. Amitalux is never the cheapest ($4.50 SchoolAI and $0 assistants are cheaper) and never the most expensive — it is the mid-priced whole platform, exactly where a pilot-stage vendor that must prove itself, but cannot afford to look like a discount tool, should sit.
Open questions — what could not be verified
- Amitalux stage: mission (“AWS staging verified with fictional data”) vs website (“local prototype”) — not reconciled. Do not quote a live-product or compliance claim to a buyer until clarified in writing.
- No Amitalux price exists and no Library Books catalogue, licensing or per-title terms were found — the $1.50 add-on is a proportional proposal only.
- Quote-only district prices could not be established for MagicSchool Enterprise, SchoolAI Pro/Scale, Brisk Premium/Intelligence, Diffit for Schools, Curipod School & District, Khan Enterprise (>1,000) and Eduaide Schools/Districts.
- Conflicts not resolved: Curipod $3,999/school (aggregator) vs custom-only; Brisk Pro $99.99/yr vs $9.99/$10/$14.99/mo; Khan district $10 (current) vs $35/$15 (older); ChatGPT for Teachers free through Jun 2027 vs Jun 2028; Brisk Common Sense 93% vs 95%; SchoolAI $4 vs $5 vs $5–$10 across secondary sources.
- Verification limits: MagicSchool, Google, SchoolAI pricing pages and the Amitalux site were read directly on 15 Nov 2026. Khan’s pricing page returned a client challenge, so $10 rests on its help centre (updated 3 Nov 2026) + secondary checks. Brisk, Diffit, Curipod, Flint and Eduaide rest on third-party checks dated May–Nov 2026.
- No efficacy evidence for Amitalux (or, rigorously, for any competitor here) was found. Do not claim learning gains, time saved, or ESSA eligibility for Amitalux.
- Gaps: no target geography specified (prices are US; Khanmigo tiers are US-only); worked-example sizes (300/600/5,000/10,000/20,000) are illustrations only; no decision yet on ESSA validation, SOC 2 audit, or Library Books contents.
Sources (as cited in the research, 15 Nov 2026)
MagicSchool — verified live: https://www.magicschool.ai/pricing
Google editions — verified live (fetch): https://edu.google.com/intl/ALL_us/workspace-for-education/editions/compare-editions/?pt_ref=pricing
SchoolAI — verified live (fetch, no prices): http://schoolai.com/pricing
Amitalux — verified live (fetch): https://www.amitalux.ai/schools · https://www.amitalux.ai/about
Khan help (updated 3 Nov 2026): https://support.khanacademy.org/hc/en-us/articles/42088874994573-How-do-I-purchase-Khan-Academy-Districts-Enterprise-Starter · https://www.khanacademy.org/schools/pricing
Linn-Mar SchoolAI contract: https://policy.linnmar.k12.ia.us/sites/policy.linnmar.k12.ia.us/files/exhibit_1006.1_schoolai.pdf
SchoolAI estimates: https://www.aiforedu.ai/tools/schoolai/ · https://techshark.io/tools/schoolai/
Brisk: https://www.learningflow.ai/compare/brisk · https://www.kuraplan.com/reviews/brisk-teaching-review · https://www.teachwithaitools.com/blog/is-brisk-teaching-worth-it
Diffit: https://www.conversiongems.com/tools/diffit · https://flintk12.com/compare/flint-vs-diffit
Curipod: https://www.toolentra.com/for/teachers/curipod-review · https://sourceforge.net/software/compare/Brisk-Teaching-vs-Curipod-vs-Gradescope-vs-GradingPal/
Flint: https://flintk12.com/compare/ai-platforms-for-school-districts · https://flintk12.com/compare/personalized-learning-platforms-for-k12
Eduaide: https://thesaninfo.com/best-ai-tools-for-teachers/ · Ed.ai: https://ed.ai/en/pricing
Buying cycle: https://civiciq.com/blog/how-to-sell-to-school-districts-the-complete-b2g-guide-for-edtech-vendors-2026 · https://vibe.us/blog/complete-guide-classroom-technology/