ChatGPT can turn a blank planning period into a working lesson draft in a couple of minutes, which is exactly why so many teachers reach for it and also why the results vary so widely.
The tool is only as good as the workflow around it. Used carelessly, it produces generic, sometimes inaccurate plans that need as much fixing as writing from scratch would have.
Used deliberately, it becomes a fast first-draft engine that frees you up for the parts of planning that actually require a teacher: judgment, context, and knowing your students.
The five approaches below move from the simplest (one strong prompt) to the most systematized (custom GPTs that already know your course).
They are not competing methods so much as a progression. Most teachers start with the first and gradually adopt the others as the habit sticks. Running through all of them is one principle worth stating early: generate, then verify, then adapt. ChatGPT drafts; you check the facts and shape the plan to your classroom.
Key takeaways
- Way 1 is a single, detailed prompt that produces a full lesson skeleton. It works best when you specify grade, subject, length, and class profile up front.
- Way 2 maps the plan onto an instructional framework like 5E, so the structure is pedagogically sound rather than just tidy.
- Way 3 starts from your actual standards and differentiates the same concept for struggling, on-level, and advanced learners.
- Way 4 treats ChatGPT as a drafting assistant inside a repeatable generate, verify, adapt routine, with fact-checking built in.
- Way 5 systematizes everything through custom GPTs and built-in tools so you stop re-explaining your context every time.
- None of these replaces teacher judgment. AI-generated content always needs a verification pass before it reaches students.
1. The single powerful prompt
This is the entry point for almost everyone: one well-built prompt that asks for a complete lesson plan in a single pass. The difference between a useful draft and a forgettable one usually comes down to how much context you give the model before it starts writing.
Specify the constraints first
Before asking for activities, tell ChatGPT who the lesson is for and what shape it needs to take. Grade level, subject, specific topic, lesson length, and a quick note on your class all change the output meaningfully. A vague prompt gets a vague plan.
Example prompt You are an experienced middle-school science teacher. Create a 45-minute lesson plan on photosynthesis for 7th graders with mixed reading levels. Include learning objectives, a hook, guided practice, independent practice, a formative assessment, and a homework task.
Ask for the sections by name
Spelling out the parts you expect (objectives, materials, activities, assessment) keeps the model from defaulting to a thin, list-style response. If you want a specific structure, describe it. ChatGPT is far better at filling a clear template than inventing one.
Iterate instead of restarting
The first draft is a starting point, not a verdict. Rather than rewriting your prompt from scratch when something is off, follow up: “differentiate the independent practice for English learners,” “add an inquiry element to the hook,” or “rewrite the objectives to align with Bloom’s taxonomy.” Each refinement builds on what is already there.
Keep a reusable core prompt
Once you find a prompt structure that consistently gives you good bones, save it. Swap out the topic and grade each time and you have a personal template that removes most of the setup work. This is the small habit that turns ChatGPT from a novelty into part of your planning routine.
2. Framework-driven lessons with 5E and similar models
A single prompt gives you a plan. Anchoring that prompt to a recognized instructional model gives you a plan with a defensible pedagogical spine. The 5E model (Engage, Explore, Explain, Elaborate, Evaluate) is a common choice in science and social studies, but the same idea applies to project-based and inquiry frameworks.
Use ChatGPT as a structure generator
Name the framework in the prompt and ask the model to organize the lesson around its phases. You get a plan that already respects the instructional logic you teach with.
Example prompt Design a 60-minute 5E lesson (Engage, Explore, Explain, Elaborate, Evaluate) on the water cycle for 5th grade. Include a short teacher script for each phase and at least one hands-on or digital activity per phase.
Generate several options per phase
Ask for three to five alternative activities for each phase, then mix and match based on your materials, time, and group. This is faster than brainstorming cold, and it surfaces ideas you might not have reached on your own.
Build in standards and skills
A clean framework is not automatically curriculum-relevant. Add your standards (NGSS, C3, state, or local) directly into the prompt so the activities map to what you are actually accountable for, not just to a nice five-part shape.
Let it handle closure and reflection
The Evaluate phase is a natural fit for AI assistance. Ask for exit tickets, reflection prompts, or short performance checks tied to the lesson objectives. These are quick wins that often get skipped when planning time runs short.
Try this comparison: Take one topic, say climate change, and ask ChatGPT to structure it three ways: as a traditional lecture-and-practice lesson, as a 5E inquiry lesson, and as a short project. Seeing the same content in three frames makes it obvious which approach fits your goals for that unit.
3. Standards-first, differentiated design
The most useful planning workflow often runs the other direction from how people first use the tool. Instead of generating a generic plan and bolting standards on afterward, you start from the standards and let everything flow from there. ChatGPT is genuinely strong at this kind of unpacking and tiering.
Start from the standards or syllabus
Paste in the relevant standard or a slice of your curriculum and ask the model to translate it into measurable objectives and a lesson sequence. You stay in control of what gets taught; the model handles the reformatting.
Example prompt Here are three Grade 9 biology standards: [paste]. Unpack them into measurable learning objectives, then propose a two-day lesson sequence aligned to those objectives.
Generate tiered materials
Differentiation is where ChatGPT saves real time. Ask for the same concept explained and practiced at three levels, then match tasks to each.
Example prompt Explain photosynthesis at three levels: for a student who is struggling, one at grade level, and one for an advanced learner. Then design a matching task for each group.
Build accommodations into the plan
Be specific about the supports your students need. You can ask the model to embed scaffolds for English learners, supports aligned to IEP goals, read-aloud-friendly phrasing, or visual supports. The more concrete the request, the more usable the output.
Draft assessments and rubrics
Once the objectives are set, ChatGPT can produce formative quizzes, exit tickets, and analytic rubrics aligned to them. Treat these as drafts to refine rather than finished instruments, but they remove the blank-page problem.
4. The human-in-the-loop workflow
This is less a single technique than a discipline that should sit underneath all the others. The premise is simple: ChatGPT drafts, the teacher verifies, and only then does the plan reach students. Codifying this as a routine is what separates careful AI use from risky AI use.
The generate-then-verify routine
Let the model produce a first draft, then deliberately switch into reviewer mode. Check the facts, scan for anything misleading, and replace generic examples with ones that fit your students and community. The verification step is not optional housekeeping; it is the part that makes the plan trustworthy.
Fact-check and check for bias
Cross-reference any factual or scientific claims against your textbook or a trusted source before teaching them. Language models can state wrong things confidently, and they can reproduce cultural blind spots. A quick check for accuracy, inclusivity, and relevance protects both your content and your students.
Use it as a brainstorming partner
You do not have to adopt a whole plan to get value. ChatGPT is useful for generating discussion questions, hooks, analogies, or extension activities even when you build the rest of the lesson yourself. Sometimes the best use is a single good idea.
Co-create reflection and metacognition tasks
Ask for student reflection prompts, peer-assessment checklists, or metacognitive questions to attach to a lesson. These deepen learning and are easy to forget under time pressure.
Be clear about AI use and student policies
Decide what you will tell students about which parts of a lesson are AI-assisted, and set clear expectations for their own use of ChatGPT. Transparency here models the kind of responsible use you want students to practice.
| Approach | Time spent | Risk | Quality |
|---|---|---|---|
| AI-generated, used as-is | Lowest | High (unchecked facts, generic fit) | Unreliable |
| AI-generated, plus verify and adapt | Moderate | Low | Strong and classroom-ready |
The middle column is the trade most teachers will want. A few extra minutes of verification buys a large reduction in risk, which is almost always worth it.
5. Systematized AI planning with custom GPTs and built-in tools
The final step is to stop starting from zero each time. Instead of re-explaining your course, your standards, and your style in every prompt, you build that context into the tool itself. A note on terminology first: ChatGPT’s old plugin system was discontinued in 2024 and replaced by custom GPTs, the GPT Store, and built-in tools. If a guide still talks about installing “plugins,” it is out of date.
Course-specific custom GPTs
A custom GPT is a configured version of ChatGPT that you set up once. You give it instructions, upload your syllabus, standards, and preferred resources, and describe your teaching style. After that, it responds in character as your course assistant, so you can request lesson plans that align with your curriculum without re-establishing context every time. For most teachers this is the single biggest time-saver in the list.
Lesson-plan generators in the GPT Store
The GPT Store includes ready-made teaching GPTs that take a subject, grade, and topic and return a structured plan with objectives, activities, and assessments. They are worth browsing, though the quality varies, so apply the same verification habit from Way 4 before trusting any output.
Connecting plans to slides and your LMS
A finished lesson rarely stops at a plan. Some edtech tools and AI slide generators can turn a lesson outline into slides, worksheets, or modules with little extra work, and custom GPTs can connect to external tools through actions and connectors. Be aware that school data-privacy rules often govern what you can connect, so check your district’s policy before linking anything to student systems.
Data-aware lessons in STEM
ChatGPT’s built-in data analysis lets you upload a real dataset and co-design an inquiry lesson around it, so students practice content and data literacy at the same time. A class exploring local weather data or census figures gets both the subject and the analytical skill in one activity.
Everyday vs power-user: A teacher new to the tool does fine with Way 1 and a saved core prompt. A tech-savvy coach or department lead gets more from a course-specific custom GPT plus connected slide and assessment tools. You can move along that spectrum at your own pace.
Best practices and caveats
A few principles keep AI lesson planning useful rather than risky, regardless of which way you use:
- Verify before you teach. Treat every plan as a draft until you have checked the facts against a trusted source.
- Watch for bias and relevance. Adjust examples and framing so they fit your students’ context and represent your community fairly.
- Mind privacy. Avoid pasting student names, grades, or other identifying data into the tool, and follow your district’s policy on AI and student information.
- Set a student-use policy. Be explicit about when and how students may use ChatGPT, and model transparent AI use yourself.
- Keep the teacher in the loop. The tool drafts and accelerates; the professional judgment about what your students need stays with you.
Frequently asked questions
Can ChatGPT write a full lesson plan in one go?
Yes, with a detailed enough prompt. If you specify grade level, subject, topic, lesson length, and a note about your class, a single prompt can produce a complete skeleton with objectives, activities, and assessment. Expect to refine it with follow-up prompts and a verification pass before it is classroom-ready.
Is it safe to rely on ChatGPT for factual content in lessons?
Not without checking. Language models can state incorrect information confidently, so any factual or scientific claim should be cross-referenced against a textbook or trusted source before you teach it. The drafting is fast; the verifying is what makes it dependable.
What is the 5E model and why pair it with ChatGPT?
The 5E model structures a lesson into five phases: Engage, Explore, Explain, Elaborate, and Evaluate. Asking ChatGPT to organize a plan around those phases produces a structure grounded in inquiry-based pedagogy rather than just a tidy outline, which is especially useful in science and social studies.
Do ChatGPT plugins still exist for teachers?
No. OpenAI discontinued the plugin system in 2024 and replaced it with custom GPTs, the GPT Store, and built-in tools such as data analysis and browsing. Guides that still reference installing plugins are out of date; look for custom GPTs and connectors instead.
How do I differentiate a lesson with ChatGPT?
Ask it to explain the same concept at multiple levels of complexity, for example for a struggling student, an on-level student, and an advanced student, and to design a matching task for each. You can also request specific accommodations, such as supports for English learners or read-aloud-friendly materials.
What is a custom GPT and is it worth setting up?
A custom GPT is a version of ChatGPT you configure once with your instructions, syllabus, standards, and teaching style. After setup, it generates plans aligned to your course without you re-explaining the context each time. For teachers who plan regularly, the upfront effort usually pays off quickly.
