Mindsmith’s Editable Video feature lets you create professional, AI-generated videos directly inside your lesson, with no external video editing software required. You describe the video you want, and the Mindsmith Agent drafts the scenes, narration, and visuals for you. From there, you can refine everything scene by scene.
There are two styles of editable video:
Animated Video is the new default. Each scene is a designed, fully animated composition: motion graphics, kinetic text, animated charts and diagrams, title cards, and narrated story beats. It’s best for explaining concepts, walking through ideas, and adding polish and movement to a lesson.
Legacy Video is the original style. Each scene is a simple layout you build from up to three tiles (a speaker, text, an image, or a video clip) placed on a scene canvas. It’s still available if you prefer the older, layout-based editor.
Both styles are narrated, previewable, and rendered to a downloadable video your learners can watch inside the lesson, and both can include an on-camera AI avatar presenter when you want one. The fastest way to create either style is to describe what you want to the Mindsmith Agent. Beyond that, the legacy editor offers a fuller set of manual controls for building and arranging each scene yourself.
Animated Video
Animated Video is the recommended way to create a video in Mindsmith. Instead of a single presenter reading a script, each scene is a designed, moving composition of animated text, images, charts, and diagrams that bring your content to life, all narrated by an AI voice and set to a consistent visual style across the whole video.
See it in action: View an example animated video
Creating an Animated Video
Animated videos are built with the Mindsmith Agent. You describe what you want, and the agent drafts the scenes for you.
Add a video tile: hover over the lesson canvas and click the Plus (+) button (or use the toolbar Video button). This inserts a blank video tile.
In the blank tile, click Create on the Editable Video card. (The same tile also offers Upload, Record, and Embed if you already have a video.)
In the Describe what video you want… box, describe your topic and what the video should cover, then generate it.
Mindsmith builds the video in the background, which usually takes a few minutes, so you can keep working on the rest of your lesson while it generates. When it’s ready, open the video to preview and refine it.
Tips for a Great Animated Video
Once you have a solid draft, these are the levers that take a video from fine to genuinely engaging:
Lead with visuals, not text. The most engaging videos lean on imagery and animated graphics rather than words on a colored background. When a scene would just be a text card, ask for a generated image, a diagram, or a chart instead. If several scenes in a row are text-on-color, you’re leaving the best of the feature on the table.
Ask for motion that matches the meaning. Rather than accepting the same gentle fade on every scene, name the treatment you want: a statistic that counts up, a comparison that grows as bars, a process that draws on step by step, a key line that types out. Motion tied to the content is what makes a point land.
Show what the words can’t, and don’t put the script on screen. Turn “our graduation rate rose eight percent” into a large animated “+8%,” not the sentence itself. Keep on-screen text to a short headline, number, or label, and let the narration do the explaining.
Time key reveals to the narration. Ask for a visual to appear right as the narrator mentions it, so the picture and the words arrive together. That timing is much of the difference between a video and a slideshow with audio.
Direct the style on purpose. Give the agent an art direction that fits the topic (muted and editorial, bold and energetic, clean and technical), and describe images the way a director would: subject, setting, lighting, mood. Left unspecified, you’ll get a generic default.
For more on writing effective prompts, see AI Prompting Tips.
Navigating the Video Editor
Click Edit on an Animated Video tile to open the fullscreen editor. It has three main areas:
The Canvas (Center): A live, playable preview of the current scene, showing your animated text, images, and graphics exactly as they’ll render.
The Sidebar (Left): Your scene list, where you manage scenes, edit each scene’s narration (transcript), and set the character and voice.
The Agent (Bottom): A prompt bar for making changes with plain-language instructions, such as “make the heading larger” or “add a bar chart comparing the two options.”
Editing Scenes
You can shape an Animated Video in two complementary ways:
Edit directly on the canvas. Click any element in the preview, such as a headline, a piece of text, or an image, to select it. You can edit text inline and adjust styling from the edit panel that appears.
Ask the Mindsmith Agent. For anything more involved, like restructuring a scene, adding a new visual, changing the animation, or adjusting the overall look, type an instruction into the agent prompt bar. Your changes populate back into the scene automatically.
Use the timeline beneath the canvas to scrub through a scene and check the pacing of the animation against the narration.
Managing Scenes and Narration
A video is made up of individual scenes, which act like slides in a presentation. You manage them in the left sidebar.
Add a Scene: Click Add Scene in the sidebar.
Reorder Scenes: Drag the scene cards to change their order.
Duplicate or Delete: Hover over a scene card, click the More Options (…) icon, and choose an action.
To set the narration for a scene, type into the transcript box on the scene’s card in the left sidebar. This is what the AI voice speaks over that scene. To keep rendering smooth, aim to keep each scene’s transcript under 600 characters.
Fine-Tuning Narration with Pauses and Pronunciations
You can control the exact pacing and delivery of your voiceover right inside the transcript:
Add a Pause: Highlight a space or word in the transcript to open the inline formatting menu and select Pause. Use the popover to set a pause of up to 3 seconds.
Adjust Pronunciation (Respelling): If the AI mispronounces a name, acronym, or technical term, highlight the word and select Pronounce, then type the phonetic spelling into the respelling field.
Choosing a Character and Voice
Even when the focus is on animated visuals, each scene’s narration uses a character voice, which you can change to match your lesson.
In the left sidebar, open the Character / Voice control for the scene.
Choose a Character from the available list. You can preview how a character looks and sounds before selecting.
Choose a Voice for that character. You can search voices and filter by language and accent.
Tip: When you change a character or voice, Mindsmith asks whether to apply the change to every scene. It’s an easy way to keep narration consistent across the whole video.
If you’d like a presenter on screen, ask the Mindsmith Agent to include an avatar in a scene. Animated videos can feature an on-camera avatar alongside your animated visuals.
Previewing Your Animated Video
Before rendering, preview your video to check the animation, layout, and audio pacing. Click the Preview button in the top toolbar to play through all scenes with narration. You can also play a single scene from the timeline.
Rendering Your Animated Video
When your scenes are ready, render the video so learners can watch it. Rendering requires organization credits, which are available on Professional and Business+ plans.
Click Render in the top-right corner of the editor.
Rendering runs in the background and typically takes around 10 minutes. You’re free to click Back to Lesson and keep working while it processes.
When rendering finishes, the video appears in your lesson and you can download the MP4.
If your video includes an avatar, you’ll also choose an avatar render engine (Avatar III or Avatar IV) before rendering. See Avatar III vs. Avatar IV for guidance on which to pick.
💡 Note: Until a video is rendered, learners won’t see it in the lesson, and the tile shows a “Not Rendered” state to you in the editor. Render (or re-render, after making changes) to publish it to learners.
Legacy Video
Legacy Video is the original editable video style. Instead of fully animated compositions, each scene is a layout you build yourself from a handful of tiles on a scene canvas. It’s still available if you prefer the older editor and its more hands-on scene controls. Like animated videos, a legacy video can include an on-camera avatar presenter.
Legacy videos use the same Editable Video tile as animated videos. What sets the legacy style apart is the editor: rather than leaning on the agent, you place and arrange each scene’s tiles yourself, choose scene templates, and fine-tune layout and animation directly, as described below.
Designing Legacy Scenes
In the legacy editor, each scene is a canvas that can hold up to three visual tiles. Use the toolbar above the canvas to build your scene:
Click Speaker, Text, Image, or Video to add a tile to the scene.
Drag the tiles on the canvas to arrange them side by side.
Click a tile to reveal its formatting options, such as text alignment on a text tile.
On a Video tile, you can upload your own short clip (up to 30 seconds) or search free stock video.
Speaker: The AI avatar that delivers the scene’s narration. Choose a character and assign a voice.
Text: On-screen labels, titles, or bullet points.
Image: Upload your own visuals or generate them with AI.
Video: Layer in short demonstration footage or B-roll.
Backgrounds and Animations
Click the scene background to open the Scene Options menu.
Use the Background color picker to set a base color. Text color adjusts automatically to stay readable.
Upload a Background Image to sit behind your tiles, and use the slider to adjust its opacity.
Use the Animation Type dropdown to add motion to the scene or an individual tile (for example, Zoom In or Slide Right), and use the slider to control the animation distance.
Rendering a Legacy Video
Click Render in the editor.
If your video includes an avatar, choose your avatar render engine:
Avatar III is a set of standard avatars that look great in most settings, and it’s the fastest to render.
Avatar IV offers higher-fidelity avatars, best for prominently featured characters.
Click Start Rendering. The estimated credit cost updates to reflect your choice. Rendering typically takes around 10 minutes and runs in the background.
For a fuller comparison of the two engines and when to use each, see Avatar III vs. Avatar IV.
Video Playback Settings
Once your video is rendered, learners can play it within the lesson, and you control how they interact with the player. Click the rendered video tile in the standard lesson editor to open its settings. You can toggle Video must be completed to require learners to watch the whole clip, and disable Learners can skip ahead to prevent them from fast-forwarding.
Related Articles:
Video is about uploading, embedding, and recording standard video
Avatar III vs. Avatar IV covers choosing a render engine when your video includes an avatar
AI Prompting Tips explains how to write effective prompts for AI generation
Media Tiles covers all image, audio, embed, and video tile options