New MLX LM Server From Apple

In three linesApple releases MLX LM Server, an inference server optimized for M5 chips. Features neural accelerator-boosted prompt processing, continuous batching for concurrent requests, and distributed inference across multiple Macs via Thunderbolt RDMA. Installable via pip.

Summary generated by Claude — human-verified