Turn web pages and articles into editable vertical videos, narrated in Portuguese and ready for review before rendering.

AI researches the source and assembles the production. You keep editorial control, open each composition, and decide when to generate the final file.
Paste a URL into the interface or pass it through the terminal. The system creates a local project in output/content2video/.
The script, storyboard, voice, captions, scenes, media, and timing remain accessible in HyperFrames Studio.
The MP4 is generated only after validation, with a button in the interface or a command in the terminal.
The default keeps the same Brazilian Portuguese voice even when the original source is in another language.
Use subscription OAuth through Codex CLI, the default, or configure an API key only in the .env.
Runs the interface and CLI.
node --version
Reuses the local OAuth login.
codex login codex login status
Encodes and inspects the final video.
ffmpeg -version
In both modes, defaults come from the .env: 60 seconds, 30% tolerance, score 8, Francisca voice, and 9:16 video.
Clone the project and create your local configuration.
git clone https://github.com/inematds/content2video.git
cd content2video
cp .env.example .env
npm installAdjust the IPs for .env if needed, and open the displayed address.
npm start # default: http://192.168.2.99:3080
The optional slug sets the name of the video folder.
npm run video:create -- https://www.inema.club/ inema-club-promoOpen the editor or run the automated check.
npm run video:preview -- inema-club-promo
npm run video:check -- inema-club-promoThe final file goes to output/content2video/<slug>/renders/<slug>.mp4.
npm run video:render -- inema-club-promoLocal production includes an INEMA.club institutional video, an English article narrated in Brazilian Portuguese, and the original MVP.


The operational foundation is ready for testing on a local network.