Base44 for Teachers: The Complete Guide (September 2026) — Build a School Web App by Describing It, and the Credit Trap

A primary Chinese panel head wanted a reading log: pupils sign in, record a book and three sentences of reflection, and the teacher sees the whole year group's progress. Quoted by an IT firm, that is tens of thousands of dollars and three months. With Base44 she had a working first version in an afternoon. It was only when she tried to roll it out to all 120 pupils that she discovered the billing unit is not how often she edits the app, but how often the pupils use it.

This guide sets out what Base44 is good for in September 2026, what it is not, a complete build walkthrough, and how the dual credit system works. The full set of nine tools is in the nine AI tools Hong Kong teachers keep asking about.

What this tool actually is

Base44 is a vibe-coding platform: describe what you want in plain language and it generates a complete, running web application including the interface, backend logic, database, user authentication and hosting. Wix acquired Base44 in June 2025 for roughly US$80 million in initial consideration plus earn-out payments through 2029, and it continues to operate as a distinct product.

On positioning: Base44 is an app builder, not an education product. There are no classes, no curriculum, no student safety gating, and it knows nothing about pedagogy. Where it differs from the three coding tools in this series is the assumption about you: Codex, Claude Code and Cursor assume you can code; Base44 assumes you cannot. For teachers, it is the only genuinely zero-code option here.

What it suits: one-off, lightweight school utilities that hold no sensitive data — an event sign-up page, a book recommendation wall, a parent survey, a panel resource index, an anniversary vote.

What it does not suit: anything that becomes a standing school system, stores marks or behaviour records, needs to integrate with the school MIS, or has to be maintained for years.

Key features

• Conversational building: describe the requirement in one prompt box and the front end, back end, database and hosting are generated automatically.
• Three chat modes: Build acts immediately, Discuss plans without changing anything (0.3 credits per message), and Edit lets you click elements in the preview to adjust them directly.
• Built-in backend: database, user authentication and management, cloud storage and an analytics dashboard are included by default.
• Model choice: Auto picks for you, or select Base 1 (Base44's in-house model), Sonnet 5, Opus 5, GPT-5.6 Terra, GPT-5.6 Sol, GPT 6 Astra or Fable 5.1 manually — manual selection requires Builder or above.
• Integration connectors: Slack, Google Workspace, GitHub and other common tools.
• In-app AI chatbot builder: put an assistant inside the app you built, for the people who use it.
• Visual editing: click an element and change colour, type or spacing; manual edits consume no credits.
• Version history: restore an earlier version, including sets of manual visual edits.

Full tutorial: a Primary 5 reading-log tracker

The scenario: Primary 5, 120 pupils, monthly entries recording what they read and three sentences of reflection, with the panel head seeing progress by year group and by class.

Step 1 — decide what will not go in. Before typing a word, set the rule: no real names, only class and class number (for example 5A-12). This is not excessive caution — the app's database lives on Base44's infrastructure, not the school's. For the full reasoning see on-premises versus cloud deployment for schools.

Step 2 — sign up and start. Register at base44.com; the free plan includes the core features. Type your first requirement into the prompt box on the home page. The more specific the prompt, the closer the first version lands:

Build a "Reading Log" web app for a Hong Kong primary school, with the interface in Traditional Chinese (Hong Kong usage).

Two user roles:
1. Pupil: signs in with a class number such as 5A-12, sees a list of their own entries, and can add a new one.
   Fields per entry: book title, author, date read, category (dropdown: story, science, history, biography, other),
   three sentences of reflection (max 150 characters), self-rating 1 to 5 stars.
2. Teacher: signs in to a dashboard showing total books read by the year group, a horizontal bar chart of
   average books per class, and a filterable table of entries by class with CSV export.

Rules:
- Do not collect real names, emails or phone numbers, only the class number.
- A pupil can see only their own entries, never a classmate's.
- Keep the interface simple enough for a 10-year-old, with large buttons.
- It must work in a mobile browser.

Step 3 — think it through in Discuss mode. Press Ctrl + . (Cmd + . on macOS) to switch to Discuss mode and ask what you have not settled:

If one pupil logs 40 entries a year, 120 pupils produce 4,800 records.
Is that a problem for the database, and will the teacher dashboard slow down?
Also, at the end of the school year I want to archive old entries without deleting them. How should that be designed?

Discuss mode costs 0.3 credits per message, while one wrong attempt in Build mode can cost several. Talking before building is the most practical way to stretch a free allowance.

Step 4 — change one thing at a time. Back in Build mode, keep each request small:

Add a "Pupils with no entry this month" list to the teacher dashboard, grouped by class, showing class numbers only.

Ask for ten things at once and the AI typically gets seven right and three wrong, and you will spend more credits undoing the three.

Step 5 — use Edit mode for appearance. Click Edit at the top of the editor and adjust type size, colour and spacing directly on the preview. Manual edits cost no credits, which makes this the cheapest way to get the layout right. For Primary 5: large type, large buttons, strong contrast.

Step 6 — be a pupil for five minutes. Walk the whole flow with a test account: sign in, add an entry, view your list, sign out. Check one thing in particular — sign in as a different class number and see whether the previous pupil's entries are visible. Broken permissions are the single most common fault in self-built apps, and the consequence is a privacy incident.

Step 7 — publish. Free-plan apps sit on a Base44 subdomain; connecting your school's own domain requires Starter or above. Have your IT coordinator review it before launch and fold the app into the school's IT usage policy.

Step 8 — pilot one class, not the whole year group. Run 5A for a month and watch how fast integration credits drain before rolling out further. That is the subject of the next section.

Pros and cons

Pros

• Genuinely usable with zero coding knowledge, the only tool in this series that can claim it.
• Idea to usable first version is typically tens of minutes rather than months.
• Backend, database, authentication and hosting are all included, so no teacher touches a server.
• Discuss mode encourages thinking before building and is markedly cheaper in credits.
• Manual visual edits cost no credits, so refining the layout carries no cost pressure.
• Base44 states explicitly that you own your code and data, with GitHub sync available.

Cons

• The dual credit system is confusing: message credits for building, integration credits for use, and the latter scale with your audience.
• Free credits carry no guarantee — the terms state that trial credits are promotional and may be modified, reduced, suspended or discontinued at Base44's sole discretion.
• Application data sits on third-party infrastructure with the school not acting as data controller.
• Choosing the AI model manually requires Builder (US$40 a month annual) or above.
• No education features and no student safety gating; every safeguard is the teacher's to think of.
• Nobody maintains a self-built app. When the teacher who made it transfers or retires, it becomes the system nobody dares touch.

Pricing and fee breakdown

Figures from the official pricing page as of September 2026, in US dollars, exclusive of tax, on annual billing.

Base44 states that annual billing saves 20% against monthly. On that basis the monthly prices work out at roughly US$20, US$50, US$100 and US$200 — check the pricing page for the exact figures. Base44 also says you can top up when credits run out, and that third-party services you connect (Stripe, SendGrid, OpenAI and so on) bill you directly.

Three things a school should price in:

First, integration credits scale with your users, not your workload. This is the one most often underestimated. You spend a few dozen message credits building the app; after that, every pupil triggering an integration action burns integration credits. The free plan's 100 a month could be gone on day one of a 120-pupil reading log. That is exactly why step 8 above pilots one class.

Second, the free tier carries no promise. The terms are blunt: free and trial credits are promotional, and Base44 may modify, reduce, suspend or discontinue them at any time and at its sole discretion, without liability. A school that builds a live system on free credits has built a school process on a promise that can be withdrawn.

Third, there is no seat multiplication — but there is no whole-school design either. Base44 pricing is unrelated to staff headcount, which is good news: 100 staff do not become 100 subscriptions. The flip side is that it was never designed for "800 pupils use this daily". A class-sized utility and a whole-school system are entirely different propositions under this pricing model.

Student accounts and age. The terms require account holders to be at least 13, or 16 in the EU, and state the platform is not intended for under-13s. Primary pupils cannot open their own Base44 account to build apps. Whether they may use an app you built is a separate layer — the data is with a third party, so the school needs its own privacy assessment.

Education offers. The official pricing FAQ says education and non-profit pricing is available case by case, verified with a student email or by contacting support. No rate is published, so ask before you budget.

Compared with the alternatives

Side by side with Edor.ai

Plainly: if a school needs a genuinely unique small utility that no product on the market provides, Base44 is far faster and far cheaper than Edor.ai, and we do not offer that capability at all. The questions below are simply outside what Base44 was designed to answer.

Why a school still needs a school-based platform

Base44 solves the problem of "there is no off-the-shelf tool for this small need", and it solves it elegantly. But nine tenths of a school's AI needs are not that problem.

It has no concept of student safety. If the app you build includes an AI chat feature, nothing gates it, nothing filters unsuitable content, nothing handles a pupil raising self-harm, and nothing leaves a record for a teacher to review. Every one of those safeguards must be imagined and implemented by the builder — and a Chinese panel head has no business needing to know any of it. In Edor.ai the four-layer guard and teacher monitoring are the default state, not an option.

It does not know your school. Your scheme of work, your panel's rubric, your notice format: each new app starts from nothing. For how a school knowledge base fixes that, see a school knowledge base that makes AI cite your own materials and policies.

Self-built apps have no successor. The most realistic risk in school IT is not security but staffing. When the enthusiastic teacher moves on, nobody knows how to change the app, where the credentials are, or how to get the data back. A vendor-maintained platform with export guarantees and support exists precisely to answer that.

It will not evidence your grant outcomes. The "智" 啟學教 funding requires interim and final reporting, and the Edor.ai PIE module produces the evidence straight from platform data — see how to use the EDB HK$500,000 AI grant.

Edor.ai is built specifically for Hong Kong schools, teachers and students, with nine modules spanning lesson planning, AI tutoring, marking, differentiation, oral practice, student data analysis, administrative workload, the PIE planning cycle and a school knowledge base. That is where all our effort goes, backed by per-school pricing, school-owned data and full export at any time. None of which replaces Base44 — that reading log really is faster to build there than with us.

Takeaway

Base44 earns a place in a teacher's toolkit: when you need a small tool nobody sells, it turns "tens of thousands of dollars and three months" into an afternoon. Two rules before you use it, though. Do not put student personal data in it, and do not build a standing school system on a free credit allowance. Use Base44 for utilities and a gated, supported, exportable school platform for everything permanent, and each does the job it is good at.

For what a school platform should offer, read the 10 requirements for an education AI platform or the feature overview.