Meta Releases Muse Spark Open, 816B Open-Weight Model

Free Instant SEO Audit

Muse Spark Open model visualized as a glowing AI sphere above two Mac Studios

By NewSunSEO October 16, 2026

Meta has released the weights for Muse Spark Open, an 816-billion-parameter multimodal model designed for coding and long-horizon agentic tasks, and it can run entirely on two Apple M5 Ultra Mac Studios. The release gives developers direct access to the model on Hugging Face, Amazon SageMaker AI, Google Cloud Model Garden, and Microsoft Foundry Models, with hosted inference available across seven platforms.

What Muse Spark Open includes

Muse Spark Open is a natively multimodal model with 816 billion total parameters, 42 billion of which are active per token. It carries a 1-million-token context window and was built around coding and agentic workloads, including complex software engineering and multi-step reasoning.

The weights ship under the Muse Model License with limited commercial terms, which allows many developers to use the model at no cost. Developers can further customize the model using the PyTorch framework, including TorchTitan for additional tuning.

Optimized for Apple silicon

Using the latest MLX framework, Muse Spark Open was optimized to run on Mac Studio with M5 Ultra. Apple’s unified memory architecture lets the integrated GPU use up to 512 GB per Mac Studio, so two machines can hold the full model. The chip’s 1.2 TB/s memory bandwidth is fast enough to read the small slice each token needs, making two Mac Studios behave like one supercomputer with a terabyte of unified memory.

On-device workflows demonstrated include shipping real iOS changes to Ice Cubes, an open-source Mastodon client, with bug fixes, new features, and a full build, as well as driving DaVinci Resolve end to end, reviewing footage frame by frame to find the strongest moments and crop them vertically.

Where the model is available

API access is live through Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter, and Together AI. Model weights are downloadable on Amazon SageMaker AI, Google Cloud Model Garden, Hugging Face, and Microsoft Foundry Models.

The open-weight release ships with custom, auto-tunable Helion GPU kernels in SGLang, delivering 1.03x to 3.95x kernel-level speedups on NVIDIA and AMD GPUs, and a 3% end-to-end speedup on AMD MI350. It runs on leading inference engines including SGLang, vLLM, TensorRT-LLM, Hugging Face Transformers, Apple MLX, llama.cpp, and Unsloth, with optimizations for AMD, Apple, Intel, and NVIDIA silicon.

Safety and preparedness

The model was evaluated in line with the Meta Superintelligence Scaling Framework, covering the most severe risks, and it met the thresholds required for release. Detailed results are in the Muse Spark Open model card, with a full evaluation report available soon. Muse Spark Open also received additional safety post-training for the open-weight release.

How to get started

To run Muse Spark Open on Apple silicon, use MLX, which pulls the weights directly from Hugging Face. An MLX Adapter is available for running the model on Mac Studio. For hosted inference, the partner platforms (Together AI, Fireworks, Baseten, Modal, Groq, Nebius, and OpenRouter) all carry the model at launch. The full collection is on Hugging Face at the Meta models page.

FAQ

What is Muse Spark Open?

Muse Spark Open is a natively multimodal US open-weight model from Meta with 816 billion total parameters (42 billion active) and a 1-million-token context window, built for coding and agentic work.

Can Muse Spark Open run on a desktop?

Yes. Using the latest MLX framework, two Apple M5 Ultra Mac Studios can hold the full model across their unified memory (up to 512 GB each) and run it on-device.

Where can developers download the weights or access the API?

Weights are available on Hugging Face, Amazon SageMaker AI, Google Cloud Model Garden, and Microsoft Foundry Models. Hosted API access is available through Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter, and Together AI.