Guide for teachers
Which AI tools turn a lesson plan into a video with narration and captions?
Updated by the Vidi team
Your first 15-second video is free. No credit card required.
If you have a lesson plan and want a short video with a voice and captions, you have several options. Some tools narrate your slides. Some put an AI presenter on screen. Some, including Vidi, build animated scenes from your text. Whichever you choose, check what it makes before students see it.
This guide covers three things. First, how to prepare the plan so the video comes out well. Second, how to pick the kind of tool that fits your content. Third, how to check facts and captions. The tool details come from each vendor's own pages, so treat them as claims to test, not as reviews.
How it works
How to make a lesson video with Vidi
Describe the class, or paste your script
Type the topic, the year level, what students already know and how long you want the video. You can also paste a student-facing script or your notes. In the chat you can paste up to 2,000 characters per message, about 300 words. Choose the narration language before you create. There are 74 to choose from.
Approve every scene's script
Vidi drafts the script scene by scene. Read each line, fix the wording and the facts, and make sure each scene carries one idea. Nothing is animated until you have checked it. Pick a look that suits the age of your class from more than 20 styles, such as clay, watercolour or hand-drawn.
Check the video, then use it
Watch it once with the sound off to check the captions. If one scene is wrong, change just that scene with a note and keep the rest. Download the video for the classroom screen, your course page or homework. Free videos are vertical. Widescreen 16:9 needs Pro or higher.
Rewrite the plan for students before you generate
A lesson plan is written for you: objectives, materials, timings and notes to yourself. A video needs words a student will hear. Some tools say you can paste a plan directly, and Leadde and Fliki both say this. Even so, consider turning the plan into a short student-facing script first, so you decide exactly what students hear. TapVid's own workflow pastes a script, not a plan.
Keep it small. Pick the one idea the video must land, not the whole lesson. Give each scene one idea and a line or two of speech. For an opener, try this structure: a question the class wants answered, three or four clear steps, and one takeaway at the end. For a recap, list the points in the order you taught them.
Write key vocabulary exactly as you want students to see it spelled. That makes the captions easier to check later. Leave out instructions meant for you, such as timings or group tasks, and give those live.
Pick the kind of tool that fits the content
TapVid's guide sorts these tools into three kinds, which is a useful way to choose. Animation tools build motion from text. Avatar tools put a presenter on screen with a voice. Template tools drop your text into ready-made scenes.
TapVid pitches animation for processes and abstract ideas, such as a chemical reaction or a supply-and-demand curve. Vidi and TapVid make this kind of video. TapVid suggests avatars for lecture-style delivery, when you want a face on screen. HeyGen and Leadde offer AI presenters, and Fliki and Pictory list them as options.
TapVid suggests template tools for quickly reusing slides. If you mainly want to reuse slides you already have, a tool built around slides may suit you best. Fliki says it narrates every slide of a PPT or PDF, and Pictory offers PPT-to-video.
Also think about where the video will play. A vertical clip fits a phone for homework. A widescreen video fits the classroom screen.
Check the draft, then check the captions
Treat whatever the tool gives you as a first draft. HeyGen calls its output a video draft, and Leadde calls it an editable first draft. TapVid's guide also talks about generating a draft and then fixing it. Some tools write the explanation for you from a topic; Fliki and InVideo describe this. Then the facts are only as good as your check.
If the tool lets you, read the script before anything is rendered. Then watch the finished video once with the sound off. TapVid's guide says auto-captions still miss technical terms and names. It also calls captions the baseline for accessibility, not an extra. Look closely at subject words, people's names and numbers.
TapVid also advises fixing the two things that bug you most rather than chasing perfection. That works for small issues of style, but fix every factual error.
If accessibility matters for your class, ask each tool whether you can download captions as a separate file or only get them burned into the picture. Fliki describes burn-in captions; we did not check whether it also offers a separate file.
Common mistakes to avoid
Showing the first draft without watching it all the way through. Watch every second before students do.
Pasting the whole lesson plan in one go. You may get a video that tries to cover everything and lands nothing. Cut it to one idea.
Trusting captions on subject vocabulary. Read them against your script.
Choosing a tool from its sample videos alone. Check whether your plan adds a watermark or caps the length. For example, Vidi's free and Starter videos carry a small logo. We did not find limits for the other tools.
Putting student names or personal details into the script. We found nothing on how these tools handle student data, so check your school's rules first.
Translating without a check. Fliki says it can localise into 80+ languages and Leadde lists 88. If you share a translated version, ask a fluent speaker to watch it first.
What each tool says it does (from vendor pages)
| What you give it | What you get | Voices and languages | |
|---|---|---|---|
| Vidi | A topic and class details, pasted notes, or a script | Animated scenes in 20+ looks, with narration and captions; up to 5 minutes on paid plans | Narration in 74 languages |
| Pictory | Notes, slides, a PDF or a Word file | Narrated, captioned video; PPT-to-video with optional AI avatars | 170+ AI voices, or record your own |
| Fliki | Lesson plans, outlines, a topic, PPT or PDF | Scene narration, auto-paired visuals, burn-in captions, quiz checkpoints, optional AI instructor | 2,000+ voices; captions and transcripts in 80+ languages |
| HeyGen | Notes, scripts, PDFs or slides | AI avatar video with voiceover and auto-synced captions | Not in our notes |
| InVideo | A topic, script or lesson outline | AI-written explanation with clips, subtitles and music on an editable timeline | Not in our notes |
| Leadde | Pasted lesson plan, script or article; .pptx, .pdf, .doc, .docx or .txt up to 200 MB | Narration, visuals, subtitles and AI teachers | 88 languages |
| TapVid | Lesson points or a script | Animated video with narration and subtitles | Not in our notes |
Limits worth knowing
- This guide relies on the vendors' own descriptions of their tools. We had no independent reviews, prices, free-plan limits or privacy terms for Pictory, Fliki, HeyGen, InVideo, Leadde or TapVid. We also did not cover tools your school may already license.
- AI drafts can be wrong in any of these tools, Vidi included. Check facts, vocabulary and captions before you teach from the video.
- In Vidi you type a topic or paste notes or a script. If you want to upload slides or a document instead, Pictory, Fliki and Leadde say they accept those files and may suit you better; Fliki says it narrates every slide. If you want an on-screen AI presenter, HeyGen or Leadde are a closer fit.
- A free Vidi account makes one 15-second vertical video. Longer videos, up to 5 minutes, need a paid plan. Widescreen 16:9 for the classroom screen needs Pro or higher.
Sources
Can I paste my lesson plan straight in?
Some tools say yes. Leadde says you can paste a lesson plan, and Fliki says it converts lesson plans and outlines. Even so, try rewriting the plan as a short script a student would hear, so you decide every line. Leave out your objectives, timings and materials. In Vidi you paste notes or a script, then edit it scene by scene.
Are auto-captions accurate enough for my students?
Not without a check. TapVid's guide says auto-captions still miss technical terms and names. Read the captions once against your script, especially subject words, names and numbers. TapVid also calls captions the baseline for accessibility, so treat them as part of the lesson, not an extra.
Do I need an AI avatar teacher in the video?
No. Avatar tools such as HeyGen and Leadde put an AI presenter on screen, which TapVid suggests for lecture-style delivery. Animation tools such as Vidi and TapVid show the idea itself in moving pictures, which TapVid pitches for processes and abstract concepts. Choose based on the content and on what your students respond to.
Can I make the same video in my students' languages?
Yes, with several tools. Fliki says it offers captions and transcripts in 80+ languages, and Leadde lists 88 languages. Vidi narrates in 74 languages and writes the script in the language you choose. Ask a fluent speaker to watch the result before you share it with the class.
How long should a lesson video be?
Short enough to hold one idea. Aim for about 60 seconds to open a lesson, or two to three minutes for a recap. If your script needs more than five or six scenes, consider splitting it into two videos. Play the video, then teach the detail yourself.
Are any of these tools free?
Fliki says it has a free forever plan, but we don't know its limits. Vidi gives a new account 5 credits and one free 15-second vertical video, with no credit card. We did not find current prices or free limits for the other tools, so check each one before you commit.
Is it safe to put student information into these tools?
We found no information on how these tools handle student data. Keep student names and personal details out of your scripts. Check your school's rules, and any list of approved tools, before you use a new service.