omarchy-mlx: MLX on Apple silicon Linux | devrel tutorial video made with ngram

About the video

This devrel tutorial video was made with ngram to explain omarchy-mlx, a project that ports apple's mlx array framework to apple silicon macs running omarchy linux. the video covers how upstream mlx depends on metal and runs on macos only, while omarchy-mlx routes the gpu path through vulkan using mesa's honeycrisp driver, with no metal and no cpu fallback, and notes that the neural engine runs separately through core ml with its own driver. it points out that gpu and neural engine support varies by chip, with the readme listing m1, m1 max, and m2 max as tested and other chips marked untested or experimental. it walks through what the project packages: the mlx python module, omarchy-mlx-chat for local browser or terminal chat, omarchy-mlx-serve for serving, and parakeet transcription on supported hardware, along with quick-start commands like omarchy-mlx-info and the first-run model download approval. it maps the repo layout, including mlx-lock, overlay and patches, docs, serve, scripts, and tests, and closes by directing viewers to contributing.md, the hardware data guide for submitting captures from untested chips, and the install guide. built with html/css motion graphics and captioned narration across nine scenes, the video functions as a technical overview for developers evaluating the project.