AI tools for destreamed Grade 9: MTH1W and SNC1W
Destreaming removed the level, not the range. One room, one credit, and students arriving from Grade 8 at four different starting points — so "differentiate" stops being a checkbox and becomes the lesson. Most AI tools were built where students are still sorted into tracks, which is why their defaults do not fit an MTH1W or SNC1W class. Here is what actually helps, and what to ignore.
MTH1W
Mathematics, Grade 9, de-streamed (2021)
SNC1W
Science, Grade 9, de-streamed (2022)
Curriculum codes quoted from the published Ontario curriculum, checked November 2026.
The short answer
What does an AI tool have to do differently for a destreamed Grade 9 class?
Serve every starting point in one room without moving the expectation. In a destreamed MTH1W or SNC1W class the code is the same for all thirty students, so the only thing worth varying is the entry point into the task: the worked example, the scaffold, the sentence frame. An “easier version” that quietly drops B3.5 down to a Grade 7 proportion problem has lowered the destination, which is the one move destreaming forbids.
Two things make that practical: published expectation codes in a picker rather than a description typed into a prompt, and three versions of one task generated from a single instruction, sharing an exit question and one rubric row. That is what the rest of this page is for — the codes, a worked three-door example, and the honest limits of what software can fix in class composition.
Sept 2021
Grade 9 math goes de-streamed
MTH1W replaces the academic and applied Grade 9 mathematics courses. Math was destreamed first.
Sept 2022
Science follows
SNC1W replaces SNC1D and SNC1P, and all Grade 9 Applied courses are removed from the province's course list.
1 code, 1 credit
No level left to sort by
Every student in the room works toward the same expectations and earns the same Grade 9 credit.
7 + 5
Strands, not streams
MTH1W has two overarching strands and five content strands; SNC1W has one overarching strand and four.
There is no US equivalent to point at. Ontario destreamed a specific set of Grade 9 courses on a specific timetable; a tool built for NGSS and honours sections has no reason to have modelled it.
Their classes are still tracked
Most AI tools were built where students are sorted into honours, AP or general sections, so the default is a level — with an easier version available on request. Destreaming took the level away. The tool has to serve every starting point in one room, and 'make it easier' is not the same instruction.
Their standards are grade bands, not codes
US tools align to NGSS or Common Core grade-band statements. MTH1W and SNC1W are Ontario courses with strand-level codes, including an SEL strand and a mathematical-processes strand that no US framework has. Mapping one to the other is manual, lossy, and lands on you.
Levelling a text is not tiering a task
The usual 'differentiate' button shortens the reading. In a destreamed class the expectation does not get shorter — B2 on ecosystems still asks for matter cycles and energy flow. If you lower the text, you can quietly lower the ceiling too. What changes is the entry point, not the destination.
| What destreaming asks of you | What that has to look like in a tool |
|---|---|
| One learning goal, several entry points | Tiered tasks that share the same expectation code, so the class ends in the same place. |
| Grade 8 foundations that did not all land | A short diagnostic before the unit, so gaps are found in planning rather than in the quiz. |
| The same final expectation for everyone | A common exit question and one rubric row, with tiered practice feeding into it. |
| English language learners in the same room | Language scaffolds on the same content — glossaries, sentence frames — not easier content. |
| IEP accommodations without separate expectations | The accommodation changes delivery and evidence, not what is being assessed. |
| Something you can defend to a department head | The code written on the resource, so the tiering is visible and arguable. |
Published expectations from MTH1W (2021) and SNC1W (2022), and the resource we would build from each
- B3Number
apply an understanding of rational numbers, ratios, rates, percentages, and proportions, in various mathematical contexts, and to solve problems
Tiered worksheet: one proportion problem, three entry points - C2Algebra
apply coding skills to represent mathematical concepts and relationships dynamically, and to solve problems, in algebra and across the other strands
Coding task scaffolded three ways, from reading code to writing it - D2Data
apply the process of mathematical modelling, using data and mathematical concepts from other strands, to represent, analyse, make predictions, and provide insight into real-life situations
Modelling task with a shared exit ticket - E1Geometry and Measurement
demonstrate an understanding of the development and use of geometric and measurement relationships, and apply these relationships to solve problems, including problems involving real-life situations
Discussion prompt set: where the formula comes from - AA1Social-Emotional Learning
develop and explore a variety of social-emotional learning skills, in the context of learning related to the other strands, including in relation to well-being and the ability to learn mathematics
Exit ticket with a confidence rating and a next-step choice
- A1STEM Skills, Careers, and Connections
apply scientific processes and an engineering design process in their investigations to develop a conceptual understanding of the science they are learning, and apply coding skills to model scientific concepts and relationships
Lab plan plus a coding task, both scaffolded for mixed readiness - B2Sustainable Ecosystems and Climate Change (Biology)
demonstrate an understanding of the dynamic and interconnected nature of ecosystems, including how matter cycles and energy flows through ecosystems
Socratic chatbot on ecosystem feedback loops, with outcome tracking - D1Principles and Applications of Electricity (Physics)
assess social, environmental, and economic impacts of electrical energy production and consumption, and describe ways to achieve sustainable practices
Role-play chatbot: an energy project hearing - E2Space Exploration (Earth and Space Science)
demonstrate an understanding of the components, characteristics, and associated phenomena of the solar system and the universe, and the importance of the Sun to processes on Earth
Assignment with an AI-proof design check
Expectation wording is quoted from the published Ontario curriculum. Overall expectations are shown; the curriculum picker in the app includes the specific expectations beneath each one.
A worked MTH1W example: the tiering changes the entry point, never the destination
pose and solve problems involving rates, percentages, and proportions in various contexts, including contexts connected to real-life applications of data, measurement, geometry, linear relations, and financial literacy
A Grade 9 class comparing two phone plans. Same expectation code on all three versions, same rubric row, same exit question.
Needs the idea in front of them
Worked example, table started, sentence frames
Complete the cost table for both plans at 5 GB, 20 GB and 35 GB using the worked example, then answer the exit question in one sentence using the frame 'Plan ___ is cheaper because ___.'
Has the procedure, needs the reasoning
The two plans and nothing else
Graph both plans, state each rate of change, and explain where the lines cross and what that point means for a student choosing a plan.
Ready to extend
The task, plus an open design brief
Design a third plan that is cheapest at 35 GB, write both equations, and defend the pricing with the graph and a two-sentence recommendation.
The door everyone leaves through
Which plan is cheaper at 35 GB, and how do you know? Same question, same rubric row, for every student in the room — only the entry point moved.
- A lesson plan with one shared learning goal and three tiered task versions
- A worksheet in three readiness tiers, each with its answer key
- A quiz whose items sit at different entry points on the same expectation
- An exit ticket the whole class answers, so the evidence is comparable
- A rubric keyed to the expectation code, not to the tier a student worked from
- A Socratic chatbot that scaffolds with questions instead of handing over answers
- Language scaffolds — glossary and sentence frames — written onto the same content
- The Ontario code stored on every record, so the tiering is visible months later
Software does not fix class composition
The hard parts of destreaming are time, class size and the spread of preparation in the room. A generator makes tiering cheaper; it does not make it free, and it does not give you another period to teach in.
Tiering is a professional judgement
The model can draft three versions of a task. It cannot know which student reads two grades below level, which one is on an IEP, or which one will shut down if the page looks like a worksheet from Grade 7. You assign the doors.
Accommodations belong to the IEP
An AI should never invent an accommodation. Use the IEP as written, and use the tool for the tiering around it — the two are not the same decision and should not be merged.
Codes only help if they are the real codes
An alignment you cannot check is worse than none. The picker holds published MTH1W and SNC1W expectations, and the code on the resource is the one from the curriculum document, so a department head can look it up.
Build the three versions once
Describe your MTH1W or SNC1W unit, attach the expectation code, and get the tiered versions with one shared exit question. The free plan is permanent and needs no card.