[KeplerBoy]

Some AI apps. You can't really load a capable LLM in 4 GB. Or does this limit not apply when dealing with WASM and WebGPU?