MAX Mini Apps: what to check before entering Russia's national messenger

By March 2026 MAX reported 107 million registered users and roughly 77 million active every day. The messenger was launched by VK in 2025 as a national platform. One window holds messaging, calls, payments, government services and mini apps.

For a business the numbers themselves matter less than what follows from them. The audience is large, and competition inside the platform is still noticeably lower than in the channels everyone already uses.

What a mini app is here

Technically the same as elsewhere. A web application opens inside the platform and receives identity, payments and a notification channel through a bot. Nothing new. The practical difference from a website is familiar too: the user installs nothing and registers nowhere, so the path to the first action is shorter. The limitation is familiar as well. A mini app is not indexed by search engines and creates no flow of people, so without an audience inside the ecosystem you end up with a working product and no visitors.

How the platform differs from Telegram

Market first. MAX targets a Russian audience and embeds services international messengers do not carry: paying state duties and fines, transfers through the national instant payment system, digital documents. For a local consumer product that moves closer to doing everything in one place, and the person never leaves for a browser halfway through.

Platform rules second. There is moderation and a partner platform for onboarding, and since 15 June 2026 self employed individuals can connect channels, bots and mini apps (previously a company or sole proprietor was required). Then a purely practical thing: MAX has fewer worked examples and fewer libraries. That is a schedule risk. It is also the reason competition is low, and everyone decides for themselves which weighs more.

Who should consider it

Companies that already have a community and client conversations inside the VK ecosystem. Then the mini app converts an existing contact instead of hunting for a new one.

Teams with a short repeating scenario: order, booking, renewal, payment. The format lives on short paths. On long ones it falls apart faster than an ordinary website. And teams already building for Telegram, because the second platform costs less than the first if the split between shared logic and the platform layer is designed in from the start.

What we built on this platform

For a service called DevOnMAX we built a corporate website: two enquiry forms, notifications to managers in Telegram, automated deployment. The service itself builds bots and mini apps for MAX, and its incoming requests show what businesses actually ask for. Most often a catalogue with ordering or a booking flow. Complex products are rare.

Our MOST. platform supports launching a store in both Telegram and MAX from one codebase. The most useful part for budgeting is this. Frontend and business logic turn out to be shared, while authentication, payments and the platform SDK differ. Design that split up front and the second platform costs noticeably less. Skip it and you get two nearly identical codebases to fix twice, with the second one always lagging behind.

What to verify before budgeting

Your own audience’s activity inside this ecosystem, not market averages. A hundred people who respond are worth more than a hundred thousand on a list.

The first scenario end to end, from catalogue to order status, and nothing beyond it. The rest belongs to version two. Analytics events tagged before launch. Without them you cannot separate people who came from your own channel from those who found the app inside the platform, and you will not know what worked.

What I do not know about MAX

We have no mini app of our own running in MAX for half a year. We built a website for a service that does this work and we keep platform support in our codebase, but a long operating history is not something I have, and passing somebody else’s numbers off as our experience is not going to happen here.

One last thing worth doing before you start, and it sounds unpleasant. At what numbers will you call the launch a failure? Fewer than a certain number of orders in the first month, for example. A threshold written down in advance removes the temptation to explain a weak result by saying you just need to wait a little longer. If you already have such a threshold, send it to us: we are curious how other people phrase it, because our own examples are still few.