Back to News

Meta releases open-weight Muse Glimmer, plans open Muse Spark 1.2

#meta#open-weights#muse-glimmer#ai-model

Meta announced the release of Muse Glimmer, a new open-weight AI model with 30 billion parameters that runs on a single GPU, under the Apache 2.0 license. The company also said it will release the weights of its more advanced Muse Spark 1.2 model in the coming weeks, marking a return to its open AI strategy after a 16-month hiatus.

Coverage timeline

  1. Hugging Face Blog
    Meta is back with Muse Glimmer: local, agentic, multimodal, and open source
  2. Techmeme
    Meta releases Muse Glimmer, a new open-weight model, and will release an open-weight version of its most advanced model, Muse Spark 1.2, in the coming weeks (Meghan Bobrowsky/Wall Street Journal)

    Meghan Bobrowsky / Wall Street Journal : Meta releases Muse Glimmer, a new open-weight model, and will release an open-weight version of its most advanced model, Muse Spark 1.2, in the coming weeks — Mark Zuckerberg has a new game plan for winning over hearts and minds to his company's artificial-intelligence efforts: open models and an open hand.

  3. Techmeme
    Meta says Muse Glimmer has 30B parameters and is "small enough" to need only one GPU; Meta plans a $1B fund to invest in US communities near its data centers (Vlad Savov/Bloomberg)

    Vlad Savov / Bloomberg : Meta says Muse Glimmer has 30B parameters and is “small enough” to need only one GPU; Meta plans a $1B fund to invest in US communities near its data centers — Meta Platforms Inc. introduced a new AI model that's light enough to run on a single computer, allowing users to download and customize the technology.

  4. TechCrunch AIRebecca Bellan
    Meta’s new Glimmer AI model offers a hint at Zuckerberg’s personal intelligence vision

    Meta’s new open-weight Muse Glimmer model offers a glimpse of Mark Zuckerberg’s personal superintelligence vision, as well as the emerging divide between AI users can own and access.

  5. Simon Willison
    Introducing Muse Glimmer

    Introducing Muse Glimmer Meta are back in the open weights game! Muse Glimmer is a brand new 30B model under a clean Apache 2.0 license (a step up from the janky Llama licenses of old). Here's a pelican which I generated using LM Studio's 18.16 GB version of the model : I really like this size of model, because if a machine has 32 GB of RAM or more (mine has 128GB) it leaves plenty of space for running other applications at the same time. Via Hacker News Tags: ai , generative-ai , llama , local-llms , llms , meta , llm-release

  6. 机器之心机器之心
    16个月后Meta杀回开源，小扎力挺蒸馏

    编辑｜杨文、陈陈 Meta 又重回开源了！ 周一，这家市值约 1.5 万亿美元的公司发布了新开放 AI 模型 Muse Glimmer 的底层参数，开发者可以直接下载并进行修改。 Meta 还表示， 未来几周内，也将开放其更强大的 Muse Spark 1.2 的模型权重 ，这是 Meta 最新的基础模型。 对此，扎克伯格表示：Meta 一直坚定支持开源，他也为这些发布感到非常自豪。 扎克伯格还在 Meta 官网发表了一篇文章，提出要把免费且能力强大的 AI 提供给数十亿人，希望借此赋予个人更多能力，并对大型机构的权力形成制衡。 文章地址：https://www.meta.com/thefutureisforeveryone/ 在一段明显针对 Google、Anthropic 和 OpenAI 等公司的表述中，他写道：其他大多数实验室都专注于为企业、政府或其他机构打造 AI。如果这些实验室最终占据领先地位，那么权力平衡将会更偏向大型机构，而不是个人。 这些承诺意味着，Meta 正重新回到其此前用来与竞争对手区分开的开放 AI 路线。 今年早些时候，Meta 曾以安全问题为由，没有开放 Muse Spark 的底层权重。随着头部 AI 公司之间的竞争不断升级，一个越来越核心的问题也变成了：最强大的 AI 技术，到底应该在多大范围内开放。 但不管怎样，扎克伯格这次重新拥抱开放路线，评论区几乎是一片叫好。 Yann LeCun 罕见地给老东家送上了掌声。 还有网友表示「这波太对味了！扎克伯格真的杀回来了。」 「我们热爱 open AI（狗头）」 Meta 重打「开源牌」 Muse Glimmer 是 Meta 自 Llama 4 以来首个开放权重模型，也是 Meta 首个采用 Apache 2.0 许可证发布的模型。 此前的 Llama 系列使用自定义许可证，对部分商业场景设有限制。Apache 2.0 宽松得多，企业可以直接部署，也可以继续微调、开发衍生模型，再将其集成进自己的产品。 Muse Glimmer 由 Muse Spark 蒸馏而来，总参数量约 296 亿，其中包含约 18 亿参数的视觉编码器。模型支持文本和图像输入，拥有 128K 上下文，重点强化多步推理、工具调用和故障恢复， 定位是个人设备上的 Agent 底座 。 传统云端 Agent 需要持续上传文