What AI actually costs an Indian college in 2026

Most AI pricing discussions in education are conducted in dollars per seat per month, which is close to useless for an Indian institution deciding an annual budget. It is worth working the numbers through properly.

Take a representative case: an engineering college with 2,500 students and 180 faculty and staff. On per-seat licensing at list price, a mainstream assistant runs roughly ₹1,800 to ₹5,000 per user per month. Licensing the 180 staff alone, at the lower end, is about ₹47 lakh a year. Licensing all 2,680 people is comfortably over ₹7 crore. Neither figure is a rounding error, and the second is not a real option for almost any private college in the country.

That arithmetic is why per-seat pricing quietly caps adoption. Institutions license administrators, the tool never reaches the students and junior faculty whose work it would most change, and the investment produces a pilot rather than a platform. The pricing model, not the technology, decides the outcome.

Institution-wide pricing inverts the incentive. Merzal AI is priced per enrolled student for on-premise deployment — ₹1,500 a year, or ₹125 per student per month, with faculty, staff and every module included. For the same college that is ₹37.5 lakh a year, covering all 2,680 people rather than 180.

The contrast is starker in dollar terms, because it removes the arithmetic. Merzal AI is $1.50 per student per month. A mainstream per-seat assistant is $25 to $30 per user per month. That is not a discount; it is a different pricing model, and it is the reason one of them can reach an entire campus and the other cannot.

Two honest caveats belong in any comparison of this kind. First, published list pricing is not what every institution pays; Google and Microsoft both offer education tiers at substantial discounts, and a serious evaluation should ask for those numbers. Second, on-premise deployment carries server sizing that is scoped separately and is not included in a per-student figure.

The number worth arguing about is not the total. It is the cost per student per month, because that is the figure a management committee can compare against everything else it funds per student — and the figure that determines whether AI reaches the whole institution or just its offices.