How much does elearning app development cost?

Published 2026 pricing guides converge on a familiar three-band shape. IIH Global puts a basic learning app at $15,000 to $35,000, a mid-complexity platform at $35,000 to $80,000, and an advanced build with AI recommendations, live classes, and gamification at $80,000 to $250,000 or more¹². SoftTeco frames the same territory more widely, at $20,000 to $500,000 for education software overall⁷. Timelines run three to four months for a basic app, five to eight for mid-complexity, and nine or more for an enterprise platform¹.

Those bands set expectations and decide nothing, because the type of elearning product you're building moves the number more than the feature list does. Here's the vocabulary first, in plain English:

Finished hour

One hour of completed course material a learner actually sits through. It's the unit the training industry prices content in, and it's the number that matters most to an edtech budget, because a platform with ten finished hours in it costs a fraction of one with two hundred.

Instructional design

The work of turning subject knowledge into a course that teaches: learning objectives, sequencing, activities, assessments, and feedback. It's a specialist discipline separate from software development, it's billed separately, and it's the single line most first-time edtech founders leave out of the budget entirely.

LMS

Learning management system: the software that hosts courses, enrols learners, tracks progress, runs assessments, and reports on completion. You can rent one, configure an open-source one, or build your own, and that choice moves the budget more than any feature list.

LTI

Learning Tools Interoperability: the standard that lets your tool launch inside a school's existing LMS, like Canvas, Moodle, or Blackboard, with the learner already signed in. If you sell to institutions, LTI support is usually the price of entry rather than a nice extra.

SCORM and xAPI

The two packaging and tracking standards for elearning content. SCORM records completions and scores inside a browser-based LMS; xAPI records richer activity across apps, simulations, and offline learning. Corporate buyers ask for both by name, and supporting them is real engineering, not a checkbox.

WCAG conformance

Meeting the Web Content Accessibility Guidelines, currently version 2.1 at Level AA for US public entities, so learners using screen readers, keyboards, or captions can complete your course. In education it's a legal requirement rather than a preference, and it covers your video content as well as your interface.

And here are the six sub-types, with what actually sets each one's price:

1. Course marketplace ($50k to $80k MVP, $80k to $250k+ full)

   The Udemy or Coursera model: many instructors, many learners. It's a two-sided marketplace before it's an education product, so you're paying for creator onboarding, revenue splits and payouts, ratings, refunds, and moderation on top of the learning features. Published estimates for a Coursera-style mobile app land at $76,500 to $103,000. The upside: instructors supply the content, so you don't fund the curriculum.

2. Single-provider course app ($15k to $35k MVP, $35k to $80k full)

   You teach, learners buy. The cheapest platform to build in edtech, because there's one catalogue, one brand, and no marketplace machinery. It's also where content dominates the budget most completely: the software is a few weeks of work and the curriculum is a year of it. Start on a hosted platform and build custom only when your teaching model needs something no product offers.

3. K-12 classroom tool ($20k to $50k MVP, $50k to $100k+ full)

   Sold to districts, used by teachers, touched by children, which is what sets the price. Expect COPPA obligations, FERPA contract terms, roster sync with a student information system, teacher and administrator views, and a district security review before anyone signs. Published figures for school management apps run $20,000 to $100,000. The build is modest; the compliance surface isn't.

4. Corporate training LMS ($50k to $80k MVP, $150k to $300k+ full)

   The compliance-and-competence workhorse. Enterprise buyers ask for SCORM and xAPI packaging, HR system and single sign-on integration, certification tracking with audit trails, and reporting a compliance officer can hand to a regulator. Published 2026 tiers put an LMS MVP at $50,000 to $80,000 and an enterprise build at $150,000 to $300,000 and up, each integration quoted separately.

5. Live tutoring platform ($40k to $70k MVP, $100k to $250k full)

   Real-time video, a shared whiteboard, time-zone-aware scheduling, tutor payouts, and a matching mechanic. Published per-type estimates put a working tutoring platform at $40,000 to $70,000. The escalators are live video quality and concurrency, plus a marketplace layer if tutors are independent rather than employed, and recording adds a per-minute storage and delivery bill.

6. Language or skill app with gamification ($30k to $50k MVP, $80k to $200k+ full)

   The Duolingo shape: short lessons, streaks, levels, and spaced repetition. Content here is huge in volume but small per item, thousands of exercises rather than dozens of video hours, which changes the production model entirely. Gamification is commonly quoted at $5,000 to $25,000 and speech recognition at $5,000 to $30,000, inside per-type ranges of $30,000 to $140,000 depending on how much AI is involved.

A caveat worth stating plainly. These ranges come from software vendors pricing their own work, not a neutral audited survey, and the per-type figures disagree at the edges. Where they did, we widened the full-build column to the top of what the same vendors quote for advanced builds in that category¹³⁶⁸⁹. Treat it as directional 2026 pricing; your number comes from a written scope. To sanity-check a band first, use our app development cost calculator and our roundup of app development cost statistics.

We hit exactly this pattern pricing a streaming service. In our teardown of what it costs to make an app like Netflix, the software lands in a comfortable five or low six figures while Netflix spends roughly $18 billion a year on content, per Variety's reporting. The app is a rounding error beside the catalogue. Edtech is the same shape at a smaller scale: for a learning platform, the courses are the content, and they're the business.

The numbers are stark per hour rather than per project. Industry estimates place custom elearning production at 80 to 300 hours of work for every one finished hour of course, depending on how interactive it is⁴. Blue Carrot's 2026 guide puts a finished hour at roughly $10,000 for template-based modules, $15,000 for tailored ones, and $20,000 for fully experiential learning with simulations⁵. AllenComm, which builds this material for a living, is blunt about the driver: interactivity is the single largest cost variable, and a simple module can start around $5,000 while an enterprise multi-course program passes $100,000⁴.

Do the arithmetic on a modest curriculum. Twenty finished hours of tailored material at the mid rate outweighs most of the platform builds in the table above. Here are the content lines that no development quote includes:

1. Instructional design

   Turning subject expertise into a course that actually teaches: objectives, sequencing, activities, and feedback loops. It's a separate profession from software engineering, billed separately, and it's the line most first-time edtech founders discover only after the platform is finished. It's also a large slice of the 80 to 300 production hours behind every finished hour of course.

2. Filming and editing

   Video is the default format and the most expensive to produce well. Published 2026 rates put 2D animated video at around $500 per finished minute and human voiceover at $20 to $150 per finished minute, with AI voice and synthetic presenters far cheaper. Either way it's a per-minute meter on a per-hour product, so it compounds fast.

3. Assessment writing

   Quizzes, practical exercises, rubrics, and anything that leads to a certificate. Good assessment is harder to write than lesson content, because a question that can be guessed teaches nothing and a certificate that isn't defensible is worthless to an employer. If you're issuing credentials, budget for a subject matter expert, not a copywriter.

4. Accessibility of the content itself

   Captions and transcripts for every video, alt text for every diagram, readable structure in every document. Professional human captioning runs about $1 to $3 per minute in 2026, with AI-only tooling cheaper and correspondingly less accurate. It's a content cost, not a development cost, and it recurs with every lesson you publish.

5. Keeping material current

   Curricula go stale. Regulations change, the software you teach gets redesigned, and a course recorded three years ago quietly stops being worth paying for. Content refresh is an annual line item in the same way a streaming service keeps buying catalogue, and it's why a beautiful, finished platform can still lose its audience.

The practical consequence: make the content before you make the platform. Teach the course live or on a hosted platform and see whether people finish it and pay again. If they don't, no amount of software fixes it. If they do, you'll know exactly what to build, which is the whole logic of scoping to an MVP first.

Rung one is a hosted course platform. Teachable runs roughly $39 to $399 a month across its plans, with a transaction fee on the entry tier, and Thinkific spans about $49 to $199. You get hosting, a player, payments, drip scheduling, and certificates this week rather than next quarter. The trade is that your product is their product, and you don't own the software.

Rung two is configuring an existing LMS. Moodle is open-source with no licence fee, and MoodleCloud plans start around $130 a year for fifty users. Canvas and Blackboard dominate the institutional market, and if you sell into schools your learners are already inside one of them, which is why LTI support usually beats building your own portal. Cleveroad prices a ready-made LMS deployment at $15,000 to $100,000 and a SaaS LMS subscription at $50,000 to $200,000 a year at enterprise scale³.

Rung three is a fully custom build. Cleveroad's 2026 breakdown puts a custom LMS at $120,000 to $300,000 or more, with an MVP at $50,000 to $80,000 over three to four months, a mid-level system at $80,000 to $150,000, and an enterprise build at $150,000 to $300,000 and up³. AllenComm puts a bespoke LMS higher still, at $200,000 to $500,000 over twelve to twenty-four months⁴. Core LMS functionality alone is estimated at roughly 2,859 development hours³.

The honest rule: custom pays off when what makes your product better is something no configuration can express. A novel assessment model, a proprietary adaptive engine, a data layer you intend to sell against. If your differentiation is the curriculum and the community rather than the software, rent the software and spend the difference on the curriculum. It's the same build-versus-buy logic we walk through for custom software generally.

The US Department of Justice's ADA Title II rule sets WCAG 2.1 Level AA as the technical standard for the websites and mobile apps of state and local government entities, which includes public school districts and public colleges. In April 2026 the DOJ extended the compliance dates by a year, to 26 April 2027 for entities serving populations of 50,000 or more and 26 April 2028 for smaller ones, while making clear it fully anticipates enforcing at the new deadline and that existing ADA obligations and litigation risk continue meanwhile¹⁰. Federally funded institutions carry Section 508 obligations on top.

In practice that means keyboard operation of everything, screen-reader labelling, sufficient colour contrast, focus management in interactive exercises, accessible documents, and captions plus transcripts for every video you publish. That last one is a content cost that recurs forever. Professional human captioning runs about $1 to $3 a minute in 2026, with AI-only tooling far cheaper and correspondingly less accurate against an accuracy bar of roughly 99 percent¹².

The retrofit penalty is the part worth internalising. HTD Health's 2026 analysis costs building accessibility in from the start at roughly $22,000 on its example project, against $70,000 to $135,000 or more to remediate the same product after launch¹¹. That's three to six times; the headline claim in the same piece is ten. We'd rather quote the table than the headline, but the direction matches what we see across this cluster: retrofitted compliance always costs a multiple of designed-in compliance, exactly as healthcare and payments compliance behaves on mobile. Sell to a public institution and you'll also be asked for a VPAT or accessibility conformance report during procurement, which is how a deal dies at the last stage. Scope it into the original written quote, not a phase two.

A B2C self-serve product needs a sign-up flow, a payment method, a good learning experience, and a reason to come back. The engineering surface is small and the cost is dominated by content and acquisition.

A B2B institutional product needs all of that plus a procurement-grade integration surface. Single sign-on through SAML or the school's identity provider, so nobody creates a password. LTI, so your tool launches inside the Canvas, Moodle, Blackboard, or Brightspace instance the institution already runs, with the learner authenticated and grades passed back. Roster and enrolment sync with a student information system, so classes populate themselves. Administrator reporting by cohort and completion, because the buyer is not the learner. And a security review before procurement signs anything.

Each of those integrations is priced separately in every LMS quote we found. Cleveroad puts integrations at roughly $2,000 to $20,000 each, with single sign-on and SCIM at $5,000 to $12,000 and SCORM or xAPI support at $5,000 to $15,000³. Selleo notes that HR system and SSO integrations commonly add $500 to $1,500 up front plus $50 to $500 a month afterwards, and that hidden integration work and user acceptance testing can add 20 to 40 percent to an initial budget¹⁴.

The strategic read: institutional sales are slower and the build is bigger, but the contracts renew and churn is a fraction of consumer churn. Just don't discover the requirement halfway through. Adding SSO, LTI, and roster sync to a product that assumed self-serve accounts is a rearchitecture, not a feature.

Under FERPA, an edtech vendor handling student education records typically operates as a school official under the institution's direct control, which means contractual limits on how you use data, security controls you can evidence, retention and deletion you can execute, and the ability to answer an access request. Under COPPA, any service directed at children under thirteen, or that knowingly collects their data, needs notice and verifiable parental consent before collection.

Architecturally that means separating learning records from analytics and marketing infrastructure, keeping an audit trail of who accessed what, being deliberate about which third-party SDKs ship at all, and making export and deletion a real feature rather than a support ticket. The tag manager and the advertising pixel don't get access to student records, which sounds obvious and very often isn't what happens.

Budget privacy work the way you'd budget accessibility: a requirement in the original scope, reviewed by someone who has read the rules. Dev Technosys prices accessibility-focused education apps at $30,000 to $140,000 as a category of their own, a fair signal that doing this properly is a real slice of a build rather than a checkbox⁶. If you sell to districts, a clean privacy posture is a sales asset, because the person signing has read the same headlines you have.

Software maintenance is the predictable one. Budget 15 to 20 percent of build cost per year for bug fixes, operating-system and browser updates, security patches, and small improvements. Some edtech vendors quote 15 to 25 percent, which is the same number with more integrations in it¹. If you shipped mobile apps as well as web, it covers store review cycles too, which we cover in our mobile app development cost guide.

Video delivery is the second meter and it behaves like a utility bill: at scale, delivery is 70 to 85 percent of run-rate infrastructure spend. We work the per-gigabyte economics through in our streaming app development cost guide, and it all applies to a video-heavy course platform. The one edtech-specific note: your viewing is concentrated, because cohorts start together, so plan for enrolment spikes rather than a smooth monthly average.

Content upkeep is the third, and it isn't in anybody's spreadsheet. Regulations change, the software you teach gets redesigned, and a course recorded three years ago quietly stops being worth its price. Every refresh is new filming, new editing, new assessment, and new captions at $1 to $3 a minute. Add the operational lines most LMS quotes list separately: user training at $5,000 to $20,000 initially plus $2,000 to $5,000 a year, data migration at $2,000 to $30,000, and localisation at $2,000 to $8,000 per language³.

If you haven't validated the teaching yet, don't build. Run the course on a hosted platform, measure completion and repeat purchase, and use what you learn to write a scope worth quoting. Our guide to scoping an MVP is the method.

If you know what you're building, the two documents that decide your budget are a feature scope and a content plan. Price them separately and honestly. Our hub guide to how much it costs to build an app and our breakdown of how much custom software costs cover bands and pricing models in full, and our custom software development and mobile app development pages explain how we build.

And if you'd rather just get a real number, our free 48-hour build plan turns a few sentences into a written scope, a milestone breakdown, and a fixed quote, with no sales call and no obligation. When you're ready to move, start a build and we'll take it from there.