If you want to make a narrated explainer video, you can write the video code yourself or ask a coding agent, software that writes and runs code from instructions, to create it from a description. explainroo uses a coding agent to write a script and scenes, then renders an MP4 file on your computer.
Remotion is a video tool built around React, a JavaScript library for user interfaces. explainroo has an agent create code files you can edit. It is for people who want narrated explainers without setting up video rendering themselves. Its narration is English-only, and it does not edit existing camera footage or create talking avatars.
Start with the video you need
Choose explainroo for a narrated explainer, animated presentation, or product demo built from an app’s screens. Choose a code-first tool such as Remotion if you want to build the video in React and edit the code yourself. With explainroo, an agent creates the first version of the script, scenes, and setup.
Check which narration language and voice options you need, whether you must edit existing footage, and where you plan to publish. Consider the local setup, the coding agent’s cost, and the rights for content you provide or generate. explainroo’s documented workflow uses an English voice, renders videos on your computer, and supports wide, tall, portrait, and square formats.
How explainroo and Remotion create videos
With a React-based video tool, developers control the video by changing code. They also create and revise the video in code. explainroo asks a coding agent to generate two files: script.md, which contains the spoken words, and scenes.js, which draws the scenes. A scene can appear at a chosen word in the narration.
To try explainroo, give a coding agent this prompt with your subject in place of the brackets:
Make me a short explainer video about [your topic]. Use explainroo for it: clone https://github.com/vincentsch/explainroo, read its AGENTS.md and follow the steps.The agent sets up explainroo, creates and checks the video, and returns an MP4 file. explainroo runs on your computer. The project supports Claude Code, Codex, Pi, OpenCode, Gemini CLI, and other coding agents that can run shell commands.
To install explainroo yourself, start with a recent version of Node.js. You will also need ffmpeg and either Chrome or Chromium. Run git clone https://github.com/vincentsch/explainroo.git, then cd explainroo, npm install, and node bin/explainroo.js doctor --fetch. During setup, explainroo downloads the voice model and the models that align scene changes with spoken words. It saves the finished video at videos/<name>/out/video.mp4. explainroo was developed and tested on Linux; macOS and Windows have had less testing. You do not need a graphics card.
explainroo narration, scenes, and editing
explainroo includes a voice model called Kokoro, which reads the script in a choice of American and British English voices. You do not need an account or API key for narration. Whisper, a speech-recognition model, analyzes the recording and marks when each word is spoken. The scene code can use those timings to show drawings or other visuals in sync with the narration.
The browser draws each frame on an HTML canvas, a surface for graphics. Scenes can include charts, code, and screenshots you provide. Lines can look hand-drawn, and the icon library includes many symbols. explainroo also creates background music and small sound effects, then lowers the music while the voice speaks.
To revise a video, ask the coding agent to change the script or scenes. There is no drag-and-drop editor, so you need to be comfortable describing changes to an agent instead of editing on a visual timeline. explainroo provides scene stills and a contact sheet of frames. A layout check flags clipped or overlapping text, while a speech check catches mispronounced words. The agent cannot watch the finished video, so these checks help it review the output.
explainroo control, formats, and costs
With explainroo, you do not have to assemble as much video-rendering infrastructure by hand. The scenes remain in JavaScript for the agent to change. You can choose among five visual styles, paper, clean, chalk, blueprint, and midnight, by changing one setting in video.json. A pace setting adjusts the voice, pauses, and animations together.
Choose a format based on where you plan to publish. Wide video fits YouTube. Tall video fits YouTube Shorts, TikTok, and Instagram Reels, with text kept clear of areas where those apps place buttons. Instagram and LinkedIn posts suit portrait or square formats. Tall, portrait, and square videos include captions that highlight words as they are spoken.
For a product demo, tell the agent which product to show and where to find its code or website. The agent recreates the screens with the product’s colors, fonts, and button labels. It then animates a pointer as it clicks through the screens and types into fields while the narration explains them.
explainroo is a free, open-source kit under the MIT license. Optional AI illustrations through OpenRouter cost money per image. The coding agent is a separate service with its own pricing. The rest of explainroo’s process has no per-video cost. explainroo does not create AI-generated live-action footage or talking avatars, and it does not edit existing camera footage. These limits matter if your project depends on filmed material or a presenter on screen.
Key Takeaways
- Choose explainroo for narrated explainers and product demos: A coding agent writes the script and scenes, then explainroo produces an MP4 on your computer.
- Use code-first video tools when direct implementation control matters: explainroo still generates JavaScript scene code, but you revise it by asking the agent for changes.
- Check the production limits before starting: Narration is English-only, and explainroo has no drag-and-drop editor. It also does not support existing camera footage.
- Budget for the agent separately: explainroo is free and open source, while the coding agent has its own pricing and optional AI illustrations cost per image.
Conclusion
For narrated explainers and product demos, explainroo uses a coding agent to create the script, visuals, timing, and MP4 from a prompt. You can edit its scenes through code, and it supports several styles and formats.
It requires a coding agent and local setup, supports English narration, and does not handle filmed footage or talking avatars.
