Super Infinity charges nothing. Launched in 2026 by Anand Kumar—the educator behind India’s Super 30 initiative, which provided free residential coaching to economically disadvantaged IIT aspirants—the platform delivers exam-focused mathematics instruction to a global audience via YouTube and a dedicated website, using animation, storytelling, and structured course design built explicitly for examination preparation. Super 30 served thirty students per residential cohort. Super Infinity is aimed at millions. A credentialed educator building content calibrated to high-stakes exams and pricing it at zero: that is precisely what the paid preparation market had long charged for.

Super Infinity is not an isolated move. The College Board partnered with Khan Academy to deliver free Digital SAT preparation through its Bluebook app; LSAC has offered free official LSAT prep since 2018, reaching hundreds of thousands of test takers; the AAMC provides a free MCAT practice exam with embedded Khan Academy lessons covering the same content and skills. Across multiple major credentialing ecosystems, exam owners themselves are distributing aligned preparation at no cost. The premise that students must pay for expert, exam-focused instruction no longer holds universally. What remains defensibly priced is something different: precise calibration to how a specific credentialing body constructs and scores its examinations—not the subject knowledge those examinations assess.

When Distribution Dissolved the Scarcity Premium

The economics of instructional scarcity rested on a simple constraint: expertise capable of producing exam-focused preparation could only reach students through channels that cost money to build and maintain—physical classrooms, published materials, staffed coaching centers. That constraint kept supply limited relative to demand. What digital distribution dismantled was not the expertise itself but the channel cost that had kept it scarce. Once a credentialed educator can publish structured preparation content globally at near-zero marginal cost, the pricing power scarcity had granted disappears along with the scarcity.

Super Infinity makes this structural shift concrete. Super 30 was defined by its residential capacity—a fixed cohort supported through free coaching and accommodation in Patna. That model was physically bounded by infrastructure. Super Infinity removes the constraint: the same examination-focused instruction, distributed through open digital channels, now reaches audiences that residential capacity could never approach. Any paid resource competing on access to expert instruction is competing on a dimension the market has broadly made free. The relevant question—for students and for the preparation market—is what digital distribution, however far it scales, cannot replicate.

Image source

Fluency Without Preparation—The Gap Open Platforms Cannot Bridge

A student can develop genuine conceptual understanding from excellent free video and still be poorly prepared for a specific credentialing examination. The distinction isn’t instructional quality—it’s what the content is calibrated to deliver. The U.S. Institute of Education Sciences’ What Works Clearinghouse identifies familiarizing students with the format of the test as a standard component of test prep programs, alongside building subject knowledge. Kulik, Kulik, and Bangert’s 1984 meta-analysis found that taking practice forms of tests raises scores, with gains increasing alongside the number of attempts—exam-like mechanics doing work that comprehension alone does not. Published assessment guidance on construct-irrelevant variance frames the underlying structural issue: when students lack familiarity with an examination’s format or tools, their results reflect that gap rather than their actual knowledge—what researchers call construct-irrelevant variance—which is why representative practice aligned to a specific assessment’s format is recommended as a remedy.

The Autodesk AI Jobs Report documents a structurally parallel gap in a different domain. Eighty percent of students report self-teaching job-relevant skills online—including through platforms like YouTube—yet only 36% feel confident using the AI tools specific to their future careers. Ninety-two percent of organizations require or prioritize certifications; only 27% of students pursue them. The pattern holds: broad online fluency and readiness for a specific credentialing system are different competencies, and self-directed video learning closes the subject knowledge gap without reliably encoding how a credentialing body evaluates that knowledge.

Credentialed examinations, it turns out, are not straightforward windows onto their subjects—they’re measurement instruments with internal logic that subject fluency alone doesn’t decode.

Examinations Are Not Transparent Windows onto Their Subjects

The Standards for Educational and Psychological Testing—jointly published by the American Educational Research Association, American Psychological Association, and National Council on Measurement in Education—treat validity and fairness as engineered properties of an assessment, not qualities that follow automatically from subject expertise. The Joint Committee on Standards for Educational and Psychological Testing, sponsored by those three bodies, states this directly in the Standards’ chapter on fairness: “Fairness is a fundamental validity issue and requires attention throughout all stages of test development and use.” Fairness is something built into an examination—not inherited from the subject it covers.

That principle carries institutional weight because operationalizing it at scale requires specialized assessment competence—not just knowledge of the subject being tested. The Australian Council for Educational Research (ACER), an independent, not-for-profit educational research organization operating across more than 100 countries, makes the distinction concrete. ACER has coordinated Australia’s participation in major international assessments—including the Trends in International Mathematics and Science Study (TIMSS) and the Progress in International Reading Literacy Study (PIRLS)—designing, managing, and interpreting complex assessment instruments for governments and education bodies worldwide. Being expert in mathematics is not the same competence as understanding how a mathematics assessment is constructed to evaluate mathematical knowledge.

A creator publishing for a broad audience optimizes for general comprehension: topic coverage shaped by audience demand, explanations pitched for non-specialist readers. Those are different decisions from those made by someone embedded in a specific credentialing ecosystem, where topic weighting, cognitive demand levels, and question-format conventions are set by the examining body rather than by viewer preferences. Those institutional constraints are built into the content itself, and reproducing them requires assessment-ecosystem embeddedness that subject knowledge alone doesn’t supply.

The Production Distinction That Platforms Cannot Replicate

Not all study methods are created equal, and the evidence makes the hierarchy explicit. Dunlosky and colleagues’ major review in Psychological Science in the Public Interest assessed common learning techniques, rating practice testing and distributed practice as high-utility strategies across a wide range of learners and materials. Rereading and highlighting—the default for most students—rated lower in the same analysis. Systems that engineer repeated, exam-like retrieval can add value that clearer explanation alone does not.

In a short-form video market saturated with free instructional content, the relevant distinction is not who can explain a topic but who can engineer exam-relevant practice and sequencing for a particular assessment ecosystem. Revision Village, an online preparation platform for IB Diploma and IGCSE students used by more than 350,000 IB students worldwide—a level of adoption that signals embedded use across the IB preparation ecosystem rather than casual supplementary browsing—produces Key Concepts videos: five–ten-minute topic summaries across IB subjects, created by IB educators including active examiners, alongside structured exam-like practice built to replicate IB assessment conditions. That content is not built by educators who happen to know IB subjects; it’s built by educators embedded in the IB assessment ecosystem, with calibration to IB-specific vocabulary, topic weighting, and examination conventions that only that embeddedness makes possible. The complication is that evaluating which resources meet that standard requires the exact knowledge the student is still trying to acquire.

For a student deciding between free general video and exam-calibrated preparation, the practical markers are which topics get emphasized relative to how the IB actually weights them, how practice is structured to mirror real exam conditions, and what an examiner actually awards marks for. The premium isn’t in the explanation—it’s in what the content has been aimed at.

The free content landscape compounds the curation challenge. Researchers at Oakland University William Beaumont School of Medicine, publishing in JMIR Medical Education, reviewed more than 1,000 biomedical education videos on YouTube and TikTok and classified 57 as probably AI-generated and low quality—flagging nonsensical analogies, factual errors, and the ongoing perpetuation of incorrect knowledge as the core hazard. View counts and recommendation algorithms are not designed to detect that drift. In a preparation context, the problem runs deeper: subject-accurate content may still encode the wrong cognitive register or emphasis weighting for a specific examination, and distinguishing the two requires assessment-ecosystem embeddedness that open platforms are not structured to provide.

What the Free Revolution Actually Clarified

The free content revolution forced a distinction the preparation market had not previously needed to make explicit: subject expertise and assessment-ecosystem embeddedness are different competencies, and only the latter grounds a durable preparation premium. Resources competing on the first now operate on commoditized terrain; those grounded in the second occupy ground that distribution infrastructure, however scalable, cannot reach. The shift has not damaged the market for preparation so much as it has exposed which part of it was resting on temporary scarcity.

Super Infinity demonstrated that expert, exam-focused mathematics instruction at global scale can now be free—and clarified, by implication, the one thing free distribution cannot provide: calibration to the specific evaluative conventions of a particular credentialing body’s examination. That is where the market for paid preparation now lives. Revision Village’s IB-calibrated content represents it at the platform level; ACER represents it at institutional scale, through the specialized assessment competence that designing and interpreting large-scale credentialed examinations demands. The meaningful question for any student evaluating a preparation resource has therefore shifted: not whether an expert teaches it, but whether the expert knows the examination—or just the subject.