RunAnywhere

About

RunAnywhere is a research-first inference lab. We hand-write the GPU and NPU kernels that make consumer silicon fast — and open-source the SDKs, infrastructure, and console that run models on every platform.

Sources