Does Kuraplan mark student work, or just plan lessons?

Details about other products were checked on 29 July 2026. Other providers change their features and pricing, so check their site before deciding.

What does Kuraplan actually do?

Kuraplan is an AI teaching assistant built for Australian and New Zealand teachers, and its published feature list is a planning list. On its New Zealand features page it names seven things: lesson plans, worksheets, unit plans, images, an AI chat, slides, and a weekly planner. Every one of those sits before the lesson rather than after it. That is a coherent product with a clear job, which is getting a teacher from a learning objective to usable classroom material without losing a Sunday to it.

Several parts are genuinely well judged. The weekly planner lets you photograph an existing paper timetable and have the AI pull out the subjects, periods and rooms, which removes the tedious setup step most planning tools leave you to do by hand. The unit planner builds in formative checkpoints and summative assessments and aligns them to achievement objectives, so assessment is designed with the unit instead of bolted on in week six. The company states it is trained on the New Zealand Curriculum and aligned to the refreshed curriculum. Its home page claims more than 50,000 teachers in total, and teachers from more than 1,900 New Zealand schools. Those are the vendor’s own figures, and the teacher count covers both sides of the Tasman rather than New Zealand alone, but the direction of the product is not in doubt.

What the site does not claim is just as informative. Nowhere on its New Zealand features page, its unit plan page or its home page does it offer to mark a submitted piece of student work, apply a rubric to a class set, or draft comments on a student’s essay. That is not a criticism. It is a product staying inside the job it was built for. Pricing is public and modest: a free tier, a Pro plan at NZD $9 a month with a seven day trial, and a schools plan at NZD $99 per teacher per year plus GST with a thirty day trial.

Which part of your workload is the real bottleneck?

Before comparing any tools, work out which half of the teaching cycle is costing you. Planning and marking both consume evenings, but they fail differently. Planning pressure feels like a blank page on Sunday night. Marking pressure feels like a pile that grows faster than you clear it, and it arrives with a reporting deadline attached. Most teachers can name which one they dread, and that answer should drive the decision more than any feature list.

There is a practical test. Look back at the last four weeks and estimate the hours. If most of your discretionary time went into building lessons, sourcing texts, and making worksheets, slides and study guides, then a planning assistant is aimed squarely at your problem and will pay for itself quickly. If your time went into reading thirty scripts and writing thirty sets of comments, a planning tool will not touch it. Buying the wrong category is the most common way a department wastes its software budget.

The two are not interchangeable, and no honest vendor pretends otherwise. A tool that generates a unit plan has not seen a single piece of your students’ work. A tool that marks student work has no view of how you intend to teach the unit. They sit at opposite ends of the same cycle, which is why many departments end up running one of each rather than hunting for a single product that claims to do both.

What does marking senior work in New Zealand actually demand?

For senior students the marking job is shaped by NCEA. Achievement in a subject is assessed through several standards, each covering a specific set of skills and knowledge, and credits accrue as students meet them. Schools use a mix of internal and external assessment: internal assessment happens in school across the year, while external assessment is the national exams at the end of it. Results come back as N, A, M or E, standing for Not Achieved, Achieved, Merit and Excellence.

That structure puts two demands on a teacher’s marking. The first is evidence. An Achieved and a Merit have to be separable by something specific in the student’s work, not by an overall impression formed on the second read. The second is consistency, because the same standard has to mean the same thing on the thirtieth script as it did on the first. Fatigue is the enemy here. Standards drift quietly towards the bottom of a pile, and the students at the end of the alphabet pay for it.

This is also why general purpose AI tends to disappoint in the senior years. A model asked to give feedback on an essay will produce fluent, encouraging, criteria-free prose. In a standards-based system that is close to useless, because it neither justifies the grade nor tells the student which criterion they missed. Anything you use at this level has to be anchored to the specific achievement standard and its criteria, and you have to be able to defend the judgement afterwards.

What should you check before paying for any AI teaching tool?

Test it on your own material rather than the demo. Give it a real unit or a real class set, ideally one you have already done by hand, then compare. A planning tool should be judged on how much editing its output needs before it is classroom-ready. A marking tool should be judged on whether you agree with its grade and whether its comments would actually help that student improve.

Then check the boring things, because they are the ones that bite later. For anything touching assessment the teacher must remain the one who sets the grade, and you should be able to edit every comment before a student sees it. Ask where student work is stored and whether your school can get a data processing agreement, since schools plans usually offer one and individual plans often do not. If a claim matters to your decision and you cannot find it on the vendor’s own documentation, treat it as not true yet.

• Run it on a real class set or a real unit you have already done by hand, then compare it against your own judgement.
• Confirm you can edit every comment or resource before it reaches a student.
• Ask where student work is stored, whether it is used to train a model, and whether a data processing agreement is available.
• Check the total cost including GST, and whether the price is per teacher or per school.
• Verify every feature claim on the vendor's own site, not on a third-party comparison page.

How does a dedicated marking assistant fit in?

For transparency, this site is published by Jeddle, which makes a tool in the marking category, so read this section as a description of how these tools work rather than a neutral verdict on the market.

A marking assistant starts from criteria you already have. You supply the rubric, the success criteria and your comment bank, connect or upload the class set, and the tool produces a first-pass grade and a set of comments for each student. You review and edit all of it before anything reaches the class. The mark, the tone and the final decision stay with the teacher. In JeddAI that is the entire product, and there is no lesson planner inside it.

The gain is not that a machine writes faster than you do. It is that the thirtieth script is measured against the same criteria as the first, and that you begin from a structured draft instead of a blank comment box. The test worth applying is whether the output resembles expert essay marking: tied to a named criterion, specific about the evidence in the work, and clear about the next improvement. If it reads as generic praise, it is not doing the job, whoever built it.

Used alongside a planning tool the split is clean. Plan the unit and design its assessment in the planning tool, then mark the work that unit produces in the marking tool. For senior classes, point the marking tool at the specific achievement standard rather than a generic essay rubric, and check the first class set against your own marking before you trust it across a cohort.