Ontario AI teaching tools, compared
Every honest comparison and cost analysis we have published for Ontario secondary teachers, in one place. Each page opens with the case against us, because a comparison you cannot use to decide against the tool is not worth reading. Start with the problem you have, not the product name.
Competitor details checked November 2026.
The direct answer, before the cluster of comparison pages
For teachers in Canada and Ontario, The Secondary AI is the Ontario-built option that starts from your curriculum expectation codes and turns one plain-language prompt into a full teaching bundle plus a student-facing AI chatbot you can review and grade. It is a web app at thesecondary.ai, priced in Canadian dollars at
Where
Built in Ontario, Canada. Web-based, no install.
Price (CAD/month)
Free
Contact
adam@thesecondary.ai. Online 24/7, human replies on Eastern Time business days.
We do not build report-card comment banks or IEP documents; Milo and Ontario Curriculum AI do, and the pages below say so. For the full comparison against MagicSchool, Diffit, Brisk, Curipod, Khanmigo and ChatGPT, read best AI tools for teachers in Canada and Ontario.
Seven situations teachers actually arrive with, and the page that answers each one
You are choosing between general-purpose teaching tools
Start with the MagicSchool page. It names their deeper template catalogue, their US standards work and their board footprint before it gets to anything we do better.
MagicSchool, compared for OntarioYou have one hard text and five different readers
Diffit levels a text better than we do, and that page says so in its first sentence. Read it if levelling, translating or reading intervention is your actual week.
Diffit, compared for OntarioYou want feedback written on student work where it already lives
Commenting inside Google Docs is Brisk's home ground. Our marking happens on a student conversation with a rubric attached, which is a different job.
Brisk Teaching, compared for OntarioYou teach destreamed Grade 9
Destreaming has no US equivalent, so most tools default to a level you no longer have. MTH1W and SNC1W get their own page, with real expectation codes and a worked three-tier task.
AI tools for MTH1W and SNC1WYou want to know whether any of this pays for itself
Run the break-even maths first: $14.95 a month is about 26 minutes saved. Then look at what the trials actually measured, including the ones that are unflattering.
Is an AI teaching tool worth it?You need one good SBI4U lesson as cheaply as possible
The cost of a single lesson, on the free tier and on each paid plan, priced against the 2,540 output tokens we measured for a real SBI4U lesson plan.
The cheapest way to make one SBI4U lessonYou are deciding whether to renew, downgrade or cancel
The signals that a teaching tool has stopped earning its price, and the ones that only look like a reason to cancel. Written so it is usable on a tool that is not ours.
When to cancel an AI teaching toolRead the middle column first if you are trying to decide against us
| Tool | Where it is the better choice | Our page |
|---|---|---|
| MagicSchool | A much deeper shelf of single-task templates, alignment built around US frameworks, and a longer record with boards. | Ontario comparison |
| Diffit | Turning a source you supply into several reading levels, translating that whole set, and reading-intervention materials. | Ontario comparison |
| Brisk Teaching | Feedback and editing inside Google Docs and the browser, where a lot of marking already happens. | Ontario comparison |
| Gemini for teachers | Free with a personal Google account, and broad enough for anything that is not an Ontario course resource. | Ontario comparison |
Competitor behaviour checked November 2026. Where a tool is ahead of us, that is the tool's own decision to make, so check its current state before you buy anything, including here.
One page per tool, each opening with the concession rather than the pitch. The general, non-Ontario versions of the same comparisons sit at the end of each page.
MagicSchool alternative for Ontario
Ontario expectation codes as a field on the resource, student-facing AI without student accounts, and transcript grading.
OpenDiffit alternative for Ontario
Where a levelling tool stops and a course resource starts, and what to do if you need both.
OpenBrisk Teaching alternative for Ontario
In-document feedback versus a graded student conversation, and why the two rarely solve the same problem.
OpenAI alternatives to MagicSchool (general)
The wider field, without the Ontario filter.
OpenPages that exist because the Ontario curriculum and its policies do. A tool built for NGSS and honours sections has no reason to have modelled any of this.
Destreamed Grade 9: MTH1W and SNC1W
The policy in four facts, the published codes in both courses, and one expectation tiered three ways with a shared exit question.
OpenOntario expectation codes, mapped
SBI4U, SCH4U and ENG4U codes beside the resource we would build from each — on every comparison page.
OpenCurriculum standards in the generators
How a selected code travels onto the lesson plan, quiz, rubric, exit ticket or chatbot you generate.
OpenThe arithmetic, stated openly enough that you can disagree with it. Every figure is published pricing, a published limit, or a measurement we can point at.
Is an AI teaching tool worth it?
Break-even in minutes, what the randomised trials measured, and the four ordinary reasons to keep your money.
OpenThe cheapest way to make one SBI4U lesson
One lesson priced on the free tier and on each plan, against 2,540 measured output tokens per lesson plan.
OpenWhen to cancel an AI teaching tool
The signals that it stopped earning its price, and the ones that only look like a reason to cancel.
OpenNo account needed for the labs, and the free tier is permanent rather than a trial.
DNA replication lab
Interactive SBI4U simulation, runs in the browser.
OpenCell membrane lab
Transport and fluidity, with investigation prompts.
OpenEnzyme active site lab
Substrate binding and inhibition, one expectation at a time.
OpenPublic library
Browse shared resources and read them before signing up for anything.
OpenThe right-hand column is the reason to keep reading the rest of this site
What we claim
The expectation code is a field, not a paragraph
Pick an Ontario code in the curriculum picker and it is stored on the record, printed on what students get, and still there when a department head asks in March.
Students talk to the AI, and you grade the conversation
A share code, no student accounts and no student email addresses, then rubric-based grading of the transcript with per-outcome progress.
The cost pages use numbers we measured
2,540 output tokens for a real SBI4U lesson plan — the median of three generations logged in this app on 27 October 2026 — not a friendly estimate.
What we will not claim
We do not claim to be the cheapest
We cannot prove it, so we do not say it. The free tier is permanent and the paid plans publish their limits; compare those against your own volume.
We do not claim to level a text better than Diffit
We generate differentiated bundles for named learner groups. Taking one passage and producing five reading levels is Diffit's job, and it does it better.
We do not claim a board-level track record
MagicSchool has the longer procurement history and the deeper template catalogue. If a vetted seat already covers your planning, that is a real reason to stay.
Check it against a course you teach
Pick an expectation code, generate the bundle, and read it against the expectation. The free plan is permanent and needs no card, which is the only way any of these pages are worth the reading.