• Oct 21, 2026 · Local & Open Source DeepSeek open-sources DeepInfer, the engine behind its own API DeepSeek released DeepInfer under Apache 2.0. It runs V4.1-Flash on Nvidia Blackwell and Huawei Ascend, but only covers offline inference for now.
• Oct 21, 2026 · Local & Open Source Perplexity Computer can now run fully offline on Apple-silicon Macs Perplexity Computer now runs its model and agent entirely on supported Macs, offline, with local files and apps. Recommended memory is 36 GB.
• Oct 21, 2026 · Local & Open Source llm-d 1.0 ships as first stable release of open Kubernetes inference stack llm-d 1.0 turns vLLM or SGLang servers into one Kubernetes pool. Red Hat, IBM and Google lead maintenance, and Google reports big TTFT and throughput gains.