Meta 开源 816B 参数模型 Muse Spark Open，适配 Apple MLX 在 Mac 上运行
After years of rivalry, Meta and Apple join hands to deliver open-weight AI to the Mac.
Meta 在 Hugging Face 上开源了 816B 参数的 Muse Spark Open，与 Muse Spark 1.3 共享同一基座模型，并追加了安全后训练。该模型每层有 256 个专家网络、每个 token 路由到其中 8 个，因此部署需要 816B 参数级别的显存但以 42B 模型的速度运行；4-bit NVFP4 版本至少需要 700GB GPU 显存，Apple 为其 MLX 框架做了适配，两台 M5 Ultra Mac Studio 各提供最高 512GB 统一内存和 1.2TB/s 带宽即可承载。它在 MCP-Atlas 工具调用测试上以 89.2 登顶，许可允许免费商用但要求标注 Built with Muse Spark Open，12 个月营收超 2000 万美元的 API 模型销售方需另与 Meta 签约。
查看原帖Meta在Huggingface上发布了Muse Spark Open的权重，这是一个拥有816B参数的模型，专为编程和智能体工作而构建。 开源版本与Muse Spark 1.3共享相同的基础模型，并额外进行了安全后训练，以抵御在没有系统护栏的情况下作为独立模型被滥用，这逆转了Meta自4月以来首个专有、闭权重前沿模型的策略。 其62层中的每一层都包含256个专家网络，并将每个token路由通过8个专家，因此部署需要816B参数模型的内存，但以42B模型的速度运行。 这一存储成本解释了与Apple的合作，因为M5 Ultra提供1.2 TB/s的统一内存带宽，而一台Mac Studio最多可搭载512GB。 在512GB M3 Ultra上测试的3-bit量化版本占用约344 GB，在短提示下生成速度约为21 tok/s。
查看原文
Meta released the weights of Muse Spark Open, on Huggingface, an 816B-parameter model built for coding and agentic work. The open version shares the same base model as Muse Spark 1.3, with additional safety post-training to resist misuse as a standalone model without system guardrails, reversing Meta’s first proprietary, closed-weight frontier model strategy from April. Each of its 62 layers holds 256 expert networks and routes every token through 8, so deployment needs the memory of an 816B-parameter model but runs at the speed of a 42B model. That storage bill explains the Apple tie-up, because the M5 Ultra pushes 1.2 TB/s of unified memory and a Mac Studio can carry up to 512GB. A 3-bit quant tested on a 512GB M3 Ultra took about 344 GB and ran at about 21 tok/s generation on a short prompt.
正文
After years of rivalry, Meta and Apple join hands to deliver open-weight AI to the Mac.
Meta released the weights for Muse Spark Open, an 816B-parameter model that runs on 2 linked Mac Studios.
Apple tuned the release for its MLX framework. Only 42B parameters fire per token because the model routes each token to 8 of 256 experts, so every step reads a small slice of memory.
Holding the whole model is the hard part, since the 4-bit NVFP4 version needs at least 700GB of GPU memory.
Each M5 Ultra Mac Studio gives its GPU up to 512GB of unified memory, so 2 machines hold the model while 1.2TB/s of bandwidth feeds each token.
On the benchmarks, Muse Spark Open does top the MCP-Atlas tool-use test at 89.2, which suits the coding agents Meta built it for.
The license allows free commercial use but requires a “Built with Muse Spark Open” label, plus a separate Meta deal for API model sellers with over $20M in 12-month revenue.
Apple hasn’t priced the 512GB Mac Studio, and a cluster of 4 machines with 256GB each costs roughly $38,000.