Tag
Kapitsa 0.2.26, an open-source Apple Silicon inference engine, publishes M5 Max benchmarks for Qwen, Nemotron, Flint and Clover models, featuring its KFlash best-first tree verification and SSD offloading that runs a 180B MoE at 61 tok/s on a 48GB MacBook Pro.
Nace.AI released Drex Max v1, an open-weight 27B multimodal decision model with a 255-way readout head and 262K-token context that returns probability distributions over options in a single forward pass, scoring 67.52 on the public Decision Index 0.4 and running locally via llama.cpp, Ollama, and MLX on Apple Silicon.
TryMirai open-sourced uzu, an Apple Silicon inference engine, featuring Weaver + DFlash speculative decoding that makes a 9B LLM (Qwen3.5) 4x faster locally and 2.6x faster than standard speculative decoding.
The open-source project ai-phone lets your Mac take over real phone lines directly. All speech recognition, LLM inference, and speech synthesis run locally, so call data never leaves your machine. It supports real-time interruptions and allows coding agents to orchestrate calls via CLI.
MKBHD reviews Apple’s newly launched touchscreen MacBook Pro: featuring a tandem OLED display (ProLED XDR), a lighter and thinner design, a starting price of $2,900, and M5 Pro/Max chips. But the touchscreen itself doesn’t address any real pain points—the real reasons to upgrade are OLED and the substantial weight reduction.
Magnitude is an open-source Rust LLM inference engine that auto-tunes GPU kernels on your local hardware, claiming 2.7x faster decode and 1.6x faster prefill than Ollama on a MacBook, with one-click integration into agents like Claude Code and Codex.
Apple announced the new 14-inch MacBook Pro powered by the M6 chip, delivering up to 30% faster graphics, 25% faster AI workloads, on-device LLM support, Wi-Fi 7, and up to 24 hours of battery life. It will be available for pre-order October 28 with availability starting November 6.
作者在 MacBook Pro (M5 Pro, 64GB) 上用 MPS 后端运行 0.9B 的 MOSS-Transcribe-Diarize 模型，对 AMI 测试集一段 17.5 分钟、4 人会议做单次转写+说话人分离，仅用 12 分钟（约 0.68x 实时）完成，4 位说话人全部识别，说话人混淆率仅 0.08%，DER 为 15.4%，并能正确处理重叠对话。
Kapitsa, an open-source inference engine optimized for Apple Silicon, achieves 128 tok/s decode on a MacBook Pro M4 Pro 48GB running Recursive's new Flint 40B Q4 model, outperforming llama.cpp and Recursive's own engine in long-context decode and prefill benchmarks.
Nace.AI's Kapitsa 0.2 ships KFlash, a best-first tree-based speculative decoding method for Apple Silicon, benchmarked against Splash 1.4.0's DFlash2, mlx-serve, Ollama, llama.cpp and others on Qwen models across M4 Pro and M1 Pro MacBooks, showing top decode throughput (~65-201 tok/s).
During the Pinduoduo 100 Billion Subsidy Double 11 event, the Mac mini M6 base model (16GB/256GB) can drop to around ¥4,299 by stacking coupons — roughly ¥2,400 cheaper than on the official site.
PyTorch 2.15 brings major MPS backend improvements for Mac users, including FP8 support, 5-6.5× faster MoE model decoding, 8× faster softmax, and dramatic memory reductions for video VAE decoding, driven by Hugging Face's PyTorch team and community contributors.
Mac.apk 是一个开源工具，可在 Apple Silicon Mac 上以原生方式直接运行 Android 应用，无需模拟器、虚拟机或 Linux 内核，应用以普通 macOS 进程运行并利用 GPU 渲染。目前处于 Alpha 阶段，已测试 65 款应用，其中 51 款可完全运行。
Nace.ai 发布开源（Apache 2.0）Apple Silicon 优化 Metal 推理引擎 Kapitsa，专为 Qwen 和 Nemotron 模型定制内核，并在 M1 Pro、M4 Pro、M5 Max 上公布对比 MLX、llama.cpp、Ollama 等引擎的解码基准。
MacNativeGames is an open-source project that runs The Legend of Zelda: Tears of the Kingdom natively on Apple silicon Macs, delivering instant loading and ~90 FPS on an M4 MacBook Air by rewriting NVN graphics to Metal 4 and answering Switch services with a small host.
Show HN 介绍 siltop——一个为 Apple Silicon（M1–M5）打造的终端系统监控工具，无需 sudo 即可显示真实功耗、时钟频率和温度，开销仅约 1.3% 单核。
llama.cpp v0.7.0 released with support for Qwen3.8-Flash-Next GGUF models via MTP-based speculative decoding on 128GB Mac/Spark setups, showcasing efficient local LLM serving.
苹果发布新版 MLX 框架，展示如何在单台 Mac Studio 上运行 Muse Spark Open（近万亿参数）等前沿模型，并通过 RDMA over Thunderbolt 驱动在多台 Mac 间实现分布式张量/流水线并行推理，在四台 M5 Ultra 集群上达到每秒 36 tokens 以上的解码速度。
Awni Hannun 展示了两台 M5 Ultra 配合 MLX 所能达到的性能水平，凸显本地 Apple Silicon 推理能力的显著提升。
Meta 发布 Muse Spark Open，宣称是美国最强开源权重模型：816B 总参数（42B 激活）、1M token 上下文、原生多模态，面向编码与智能体任务，并首次可在两台 Apple M5 Ultra Mac Studio 桌面端本地运行。