How to Use ChatGPT to Write Lesson Plans for Free

My department head forwarded me a lesson plan last spring and asked what I thought of it.

It was clean. Professionally formatted. The objective was written in proper educator language. The activities had timing. There was a warm up, a main task, an exit ticket. On paper it looked like exactly what a lesson plan should look like.

I read it three times before I figured out what was wrong with it. The lesson had no instructional thinking underneath the structure. Every activity existed but none of them connected to each other in a way that built understanding progressively. The exit ticket assessed a different skill from the main task. The differentiation note at the bottom said "provide additional support for struggling learners" which tells a teacher absolutely nothing about what to do.

My department head told me a colleague had generated it using ChatGPT in about forty five seconds.

That story is why this guide exists. Not because ChatGPT cannot write good lesson plans. It can, and I have been using it regularly for months to do exactly that. But forty five seconds and a vague prompt produces a lesson plan shaped object that will disappoint you in the room. The right process produces something genuinely usable. The difference between the two is entirely in how you prompt and what you do with the output.

This guide gives you the right process, tested across real classrooms, with the exact prompts that work.

What the Free Tier of ChatGPT Can and Cannot Do

Before the process, the honest picture of what you are working with on the free tier.

ChatGPT free tier in 2026 gives you access to a capable language model that can generate structured lesson plans, adapt them to grade levels and subjects, incorporate specific standards when you provide them, and revise its output based on your feedback. The free tier has usage limits that you will occasionally hit during heavy planning sessions, but for most teachers planning a week of lessons it is sufficient.

What the free tier does not do automatically: align lessons to specific standards without being told the standard explicitly, differentiate in ways that are actually useful without being told the specific needs of your class, generate activity sequences that have genuine pedagogical logic connecting them without being guided to do so, or produce anything other than the most predictable version of a lesson on any topic without specific creative direction.

The free tier is a strong tool for a teacher who knows how to use it. It is a template generator for a teacher who does not. This guide puts you in the first category.

The One Thing That Changes Everything Before You Start

There is a practice that separates teachers who get genuinely useful ChatGPT lesson plans from teachers who get the structured nothing my department head forwarded me.

Before you type a single word into ChatGPT, write the answers to these four questions somewhere you can see them while you prompt.

What do my students already know that connects to this lesson?

What specific gap or misconception am I trying to address, not just what topic am I covering?

What should a student be able to do or explain at the end of this lesson that they cannot do now?

What are the real constraints: time available, materials I actually have, technology that actually works, any student needs I am planning around?

These four answers are the professional knowledge that ChatGPT does not have and cannot generate. Every good prompt is built from them. Every generic output comes from a prompt that skipped them.

Five minutes before prompting. Every time.

Step by Step: How to Use ChatGPT to Write Lesson Plans

Step One: Set the Context in Your First Message

ChatGPT responds to the context you establish at the start of a conversation. A first message that gives it the full picture produces better output across every subsequent exchange in that session than a first message that is vague and then tries to correct course through follow up prompts.

Here is the first message structure that consistently produced strong output across my testing:

"I am a teacher and I need help planning a lesson. Before you generate anything, I want to give you the context you need to produce something actually useful rather than generic.

Grade level: [your grade] Subject: [your subject] Standard or learning objective: [paste the full text of the standard, not just a code number] What students already know: [your answer from the pre prompt questions] The specific gap I am addressing: [your answer] What students should be able to do by the end: [your answer] Constraints: [time, materials, technology, any student population notes]

Do not generate the lesson plan yet. First tell me what you understand the lesson's central challenge to be, based on what I have told you. I want to check that you have understood the context correctly before we build the plan."

That last instruction is the one most teachers do not use and the one that makes the biggest difference. Asking ChatGPT to reflect back its understanding of the lesson before generating anything catches misunderstandings early. If its summary of the central challenge is wrong, correct it before the plan is built. If the summary is right, you have confirmation that the generation that follows will be grounded in your actual situation.

Time: five to seven minutes to write this message.

Step Two: Generate the Lesson Plan With Specific Structure Requirements

Once ChatGPT has confirmed its understanding of the context, ask for the plan with explicit structural requirements. Do not let it decide the structure. Tell it exactly what you need.

"Now generate the lesson plan. Include these specific components:

A warm up activity of five to eight minutes that activates relevant prior knowledge without revealing the lesson's central concept. Explain in one sentence why you chose this specific warm up.

An instructional sequence with realistic time allocations for each segment. Each segment should connect to the next in a way you explain briefly.

A student task that requires genuine thinking or application, not recall or copying. The task should produce something I can look at to understand whether students understood the lesson.

A formative assessment checkpoint, not a participation activity, that tells me specifically whether students grasped the lesson's central idea before they leave.

A closure of three to five minutes that connects today's learning to what comes next in the unit.

Differentiation built into the activity itself for two groups: students who need additional scaffolding and students ready for extension. Do not put this in a separate note at the bottom. Build it into the task description.

Include the timing for each component. The total should not exceed [your period length]."

This instruction set produces a structurally complete plan. The instruction to explain the reasoning behind each component is what makes the difference between a plan that looks right and a plan that actually is right.

Time: thirty seconds to send this message. Two to three minutes for ChatGPT to generate.

Step Three: Use the Free Tier's Conversation Feature to Refine

This is where the free tier of ChatGPT earns its place over static AI form generators. You can have a conversation. You can push back. You can ask for revisions.

After reading the initial output, identify one or two specific things that do not work for your class and ask for targeted revisions. Keep these specific.

Do not say: "Can you make the lesson more engaging?"

Do say: "The main task has students working individually and silently for twenty minutes. My class does better with structured partner work at that stage. Can you redesign the main task as a structured partner activity that still produces individual evidence of thinking?"

Do not say: "The differentiation does not feel right."

Do say: "The extension task is just more questions of the same type. Can you replace it with a task that requires students to apply the concept in a different context or explain it to an imaginary student who is confused?"

Specific revision requests produce specific improvements. Vague revision requests produce a slightly different version of the same problem.

Each revision round takes two to three minutes. Most lessons need one or two rounds. By the third revision you typically have something genuinely usable.

Time: five to ten minutes for revision rounds.

Step Four: Run the Output Through a Checklist Before Using It

Four things to verify before any ChatGPT lesson plan goes into your planning documents.

First, timing. Add up every time allocation in the plan. Does it fit your period? ChatGPT frequently generates plans that run over by five to ten minutes. Catch this now, not when you are three activities in and running out of time.

Second, standard alignment. Read the standard you provided in Step One. Read the lesson's main task. Does the task actually address the standard or does it address an adjacent skill? If the standard requires students to construct an argument with evidence and the main task asks them to summarise a reading, the plan is not aligned regardless of what the objective says.

Third, formative assessment validity. Look at the exit ticket or formative checkpoint. If every student could complete it by remembering a fact from last week rather than applying today's learning, it is not assessing what today's lesson taught. Change it.

Fourth, the differentiation reality check. Read the scaffolding for below level students. Is it actually supportive or is it just the same task with fewer questions? Read the extension. Is it genuinely challenging or is it just more of the same? If either description fits, go back and ask ChatGPT for a genuine revision using the specific language from Step Three.

Time: four minutes.

Step Five: Add What ChatGPT Cannot Know

ChatGPT does not know your students. It does not know that one of them asked a question last week that connects perfectly to today's lesson and that you could open with. It does not know that the example in the warm up references a context your students have no familiarity with. It does not know that the seating arrangement it assumed will not work in your room.

Read the plan once more with your specific class in mind. Make the changes that only you can make. In my own testing this produced an average of two to four changes per plan. None of them took more than a minute each. All of them made the plan meaningfully better for the actual students receiving it.

Time: five minutes.

Step Six: Save the Conversation Prompt, Not Just the Plan

When a conversation produces a strong lesson plan, copy the opening context message and the structure requirements message into a document you keep specifically for this purpose.

Label it by subject, grade level, and lesson type. Build this library over time. Within a month of consistent use you will have a bank of tested, refined starting prompts that dramatically reduce the time cost of Step One and Step Two for future lessons.

In the final weeks of my testing period, lessons that had taken me forty minutes to plan with this workflow took twenty minutes because the prompt library reduced the front end work to editing an existing prompt rather than building from scratch.

Time to save: two minutes. Return on investment: compounding across every future lesson in that subject.

The Common Mistakes and How to Avoid Them

These are the four mistakes I see most frequently when teachers use ChatGPT for lesson planning.

Mistake one: prompting with the topic and nothing else. "Write a lesson plan on the water cycle for 5th grade" is the prompt that produces the structured nothing. The four pre prompt questions are not optional. They are what makes the difference.

Mistake two: accepting the first output. ChatGPT's first generation is a starting point, not a finished product. The revision conversation in Step Three is where the plan becomes actually usable. Teachers who skip this step and use the first output are using half the process.

Mistake three: treating ChatGPT's timing as accurate. The timing in ChatGPT generated plans is frequently optimistic. Activities that would take a real class of students twenty five minutes are often allocated fifteen. Always verify timing against your knowledge of your students' pace.

Mistake four: skipping the checklist because the plan looks good. Plans that look good and teach wrong are the specific failure mode this checklist is designed to catch. The standard alignment check and the formative assessment validity check exist because ChatGPT confidently produces misalignment that reads as alignment until you look carefully.

A Complete Example: The Prompt and the Plan

Here is a real example from my testing. The subject is 8th grade English. The standard is CCSS.ELA.W.8.1: Write arguments to support claims with clear reasons and relevant evidence.

The opening context message:

"I am an 8th grade English teacher planning a lesson on argument writing. My students understand what a claim is and can write a basic thesis statement but they consistently write evidence without explaining how it connects to their claim. They drop quotes and move on. The gap I am addressing is the analysis layer of argument writing, specifically the move from evidence to explanation of relevance. By the end of the lesson I want students to be able to take a piece of evidence they have selected and write two to three sentences explaining precisely how that evidence supports their specific claim, not just what the evidence says. Constraints: 50 minutes, no technology needed, class of 26 students including six who benefit from sentence level scaffolding and four who are ready for complex analytical work. The standard is CCSS ELA W 8 1. Do not generate the lesson plan yet. Tell me what you understand the central challenge of this lesson to be."

ChatGPT's understanding check confirmed it understood the gap correctly: students are quoting without analysing. The plan it then generated included a warm up where students evaluated two versions of the same paragraph (one with bare evidence, one with analysis) and identified what the second version did differently. The main task had students work with a partner to write analysis sentences for evidence they chose from a shared text. The scaffolding for below level students was a sentence frame: "This evidence shows [claim connection] because [explanation]." The extension was to write a second analysis sentence that addressed a counterargument the evidence could raise.

Harriet from my UK review read that plan. She said it was the kind of lesson structure she would have spent an hour designing on her own.

It took twenty two minutes with the workflow.

Final Verdict

ChatGPT writes lesson plans for free and it does so well enough to be genuinely useful when the process is right. The free tier is sufficient for most teachers' weekly planning needs. The gap between a plan that wastes your time and a plan that works in the room is not the tool. It is the process.

The four pre prompt questions before you open the tab. The context and understanding check before you generate. The specific revision requests rather than vague ones. The checklist before the plan goes into your week. The five minute pass where you add what only you can add.

That is the process. It takes roughly twenty to thirty minutes per lesson once it is established, compared to the forty five to ninety minutes that good lesson planning costs from scratch. Across a week of five lessons it returns two to four hours. Across a school year it returns a number large enough to matter.

My department head's colleague had the right tool. She had the wrong process. There is a version of this story that ends differently, and this guide is it.