Students use dynamic media like videos to understand math or science concepts since the visualizations in textbooks are not enough. However, watching lengthy videos takes time. Videos cannot be quickly referred. Most teachers cannot make such engaging content are often helpless when it comes to providing high-quality animated illustrations. Hence, ManimBooks provide a hybrid solution where the scientific concepts can be animated inside the book itself using high-quality manim based illustrations. ManimBook is an E-book format with basic text formatting features like other formats (like epub) but with the benefit of embedding manim animations. They can be referred to quickly, offline or online, and the text is readily available for citation.
The project consists of mainly 3 different software as referenced by the submodules of this repository.
-
A service written in Go is responsible for compiling and previewing the manimbook chapters. It uses gin and manim to achieve desired functionality.
-
The web-based editor is made with Next.js, shadcn, blueprint and other frontend javascript libraries. It interacts with the backend service to generate previews of the current chapter.
-
Finally, there is an offline reader that can be used to view and read a compiled manimbook file.
Tip
Use our hosted editor at https://manimbook.kush.in/ide to save some trouble ;)
Instructions to use individual services/software along with their architecture are provided in their respective repositories.
You can locally deploy the entire editor stack using docker compose. An example configuration is provided here.
Download the compose.yaml in an empty folder or copy it from below
wget -O compose.yaml https://raw.githubusercontent.com/ru3bix/fosshack25/refs/heads/main/compose.yaml
services:
mbook-frontend:
build: https://github.com/ru3bix/manimbook-editor.git
environment:
- BASE_URL=http://mbook-runner:8080
- AI_KEY=<an openai compatible api key>
ports:
- 3000:3000
mbook-runner:
build: https://github.com/ru3bix/manimbook-runner.git
ports:
- 8080:8080
environment:
- BASE_DIR=/app
- PUBLIC_HTTP_URL=http://127.0.0.1:8080
- PUBLIC_WS_URL=ws://127.0.0.1:8080Edit the environment variables. Make sure to add an OpenAI compatible API key. You can get one by self-hosting OpenLLM. Start the service with
docker compose up -d
Important
Keep in mind that the AI helper will require an OpenAI compatible API key to work.
Caution
DO NOT use this to host your own service for public to access. Currently the project is in VERY early stages and the entire runner service is just one big RCE vulnerability.
To save the trouble, we have hosted the editor and runner service at https://manimbook.kush.in/ide and https://mbook-runner.kush.in respectively.