Meta Muse Spark Open lands on AWS and Macs
What you need to know
Meta released Muse Spark Open, an open-weight mixture-of-experts model with about 816B total parameters and roughly 42B active per token. AWS says the coding-optimized model is now in Amazon SageMaker JumpStart, but gave no pricing, regions, or performance details. Meta calls it the most capable US open-weight model to date and the first in its class able to run on a desktop, using two Apple M5 Ultra Mac Studios with the latest MLX framework. Meta and Apple worked together on the launch, and Apple published a technical post about MLX, Muse Spark Open, and two other open models. The model is natively multimodal and aimed at long-horizon tasks, complex software engineering, and multi-step reasoning.
Generated by AI from the reporting · updated 14 minutes ago
Developments
3 developments
- Oct 15 18:15 · 2 reportsMeta and Apple team up to make Muse Spark Open AI model runThe Verge · AI: Meta and Apple team up to make Muse Spark Open AI model run better on Macs
- Oct 15 17:43 · 1 reportsMeta's Muse Spark Open coding model arrives in Amazon SageMaAWS Machine Learning Blog: Meta's Muse Spark Open coding model arrives in Amazon SageMaker JumpStart
- Oct 15 00:00 · 2 reportsMeta releases Muse Spark Open, an open-weight 816B-parameterHugging Face Blog: Meta releases Muse Spark Open, an open-weight 816B-parameter multimodal MoE model
Timeline
Follow the reports to see every side of the story.
- TechCrunch · AIMeta launches Muse Spark Open, an open-weight model built to run on Macs
Meta launched Muse Spark Open, an open-weight model it calls the most capable in the United States.
- The DecoderFeaturedMuse Spark Open is Meta's answer to Qwen, GLM, and Kimi in the open-weight race
Meta has released Muse Spark Open, an open-weight model it calls the most capable US open-weight model.
- The Verge · AIFeaturedMeta and Apple team up to make Muse Spark Open AI model run better on Macs
Meta announced the Muse Spark Open AI model, which it calls the most capable US open-weight model to date and the first in its class able to run on a desktop using two Apple M5 Ultra Mac Studios with the latest MLX framework. The 816-billion-parameter natively multimodal model is aimed at long-horizon tasks, complex software engineering, and multi-step reasoning, and Meta and Apple worked together on the launch, with Mark Zuckerberg and new Apple CEO John Ternus publicly praising each other. Apple also published a technical post about MLX, Muse Spark Open, and two other open models.
- AWS Machine Learning BlogMeta's Muse Spark Open coding model arrives in Amazon SageMaker JumpStart
Meta's coding-optimized open-weight model Muse Spark Open is now available in Amazon SageMaker JumpStart.
- Hugging Face BlogFeaturedMeta releases Muse Spark Open, an open-weight 816B-parameter multimodal MoE model
Meta released Muse Spark Open, an open-weight mixture-of-experts model with roughly 816B total parameters and about 42B active per token.
Heat trend for this story
Not enough continuous data yet to chart a trend.