Devotd Github
NewDeveloper tools

gatewai-dev/framefields — WebGPU-based code-driven video tool

gatewai-dev/framefields is a package for building video compositing, motion graphics, and 3D scenes in TypeScript code and rendering them on WebGPU.

Repositorygatewai-dev/framefields ↗

gatewai-dev/framefields is a package for building video compositing, motion graphics, and 3D scenes in TypeScript code and rendering them on WebGPU.

GitHub description: Code-first video & image, rendered natively on WebGPU.

stars
★ 496
Forks
32
Language
TypeScript
License
Apache-2.0
Read original source ↗View on GitHub ↗

gatewai-dev/framefields is an npm package that defines videos and images in TypeScript code and renders them natively on WebGPU. According to the repository description, it is designed to let coding agents create and render video compositing, motion graphics, and 3D scenes. There is no registered latest release. So this is an introduction to the capabilities the repository offers rather than a rundown of changes in any specific version.

This package handles video compositing and visual effects, typography and motion graphics, and 3D capabilities in code. WebGPU is an API that hands rendering off to the graphics processing unit. Framefields renders without going through Chromium or Puppeteer, on native GPU paths such as Dawn/WebGPU and Metal/Vulkan. It supports deterministic per-frame rendering, and you can inspect the frames you produce before exporting to a hardware-encoded MP4. It is licensed under Apache-2.0, and the repository records 496 stars.

That said, the provided records do not present comparative data demonstrating the scope of the features or their performance. The research notes point out that the monorepo contains eight renderer packages and nine shader files, and assess that the share of the visual effects handled by GPU code is unclear. Rendering quality or speed is hard to judge from this information alone. Framefields can be viewed as a code-based media production path, but the actual performance and range of effects are better left open until separate supporting evidence is confirmed.