None
None
Long Story Video Skill
Turn your idea into 10s–10min video—AI writes scripts, prompts, and generates footage automatically.
Ads Video Skill
Generate professional ads and sales videos—AI auto-generates scripts, prompts, and footage.
3D Science Explainer Video Skill
Convert scientific concepts into stunning 3D explain animations
Feedback
😢 Every free AI generation costs us real GPU money. We pay out of pocket to keep this free for you.
💎 Upgrade to Pro and unlock:
- ✓HD Download
- ✓No Watermark
- ✓Unlimited Generations
Muse Spark Open
Muse Spark Open: open weights, 816B total / 42B active parameters, 1M-token context, and desktop inference on two Apple M5 Ultra Mac Studios.
All Tools
Discover our comprehensive AI-powered animation toolkit
MiniMax H3
MiniMax H3 AI Video Generator
Seedance 2.5
The Future of AI Video Is Here.
Seedance 2.0
The Future of AI Video Is Here.
Veo3.1
Create Stunning Videos with Veo3.1
FLUX 3 Video Generator
Kling 3.0
Next-Gen AI Video Generator
Grok Video Generator
Create Videos from Text or Images with AI
MiniMax H3 video generator
MiniMax H3 AI Video Generator
Why Run Muse Spark Open
Muse Spark Open lands as open weights: a natively multimodal model for coding, agentic work and long-horizon tasks that you can host yourself.
- Open Weights, US-BuiltMuse Spark Open arrives today as the strongest open-weight release to come out of the US — 816B total parameters, 42B active, a 1M-token window and native multimodality.
- Superintelligence on a DesktopFrontier-class models should not be locked in a datacenter; Muse Spark Open boots on your own desk, pairing a couple of Apple M5 Ultra Mac Studios with the newest MLX build.
- Agentic Work Proven On-DeviceMuse Code on Muse Spark Open shipped real iOS fixes and features to Ice Cubes, then drove DaVinci Resolve end to end, cropping the strongest frames vertically.
How to Run Muse Spark Open
Three routes to put Muse Spark Open to work — local Apple silicon, an inference engine of your choice, or a hosted API.
Muse Spark Open Capabilities
A million-token window, native multimodality, desktop-class memory bandwidth and open weights — the parts of Muse Spark Open that make agentic video work possible.
1M-Token Context Window
Muse Spark Open holds a 1M-token window, so lengthy agentic sessions and multi-step reasoning chains keep earlier work in view.
Long-Horizon Agentic Coding
Tuned for software engineering and agentic loops, Muse Spark Open handles long-horizon jobs and end-to-end control of tools such as DaVinci Resolve.
Native Multimodality
Muse Spark Open reads mixed inputs directly rather than through bolted-on adapters, which is what makes frame-by-frame footage review possible.
Two Mac Studios, One Supercomputer
Unified memory lets each M5 Ultra GPU address up to 512 GB, so a pair of Mac Studios holds the whole Muse Spark Open model as one machine with a terabyte of pooled memory.
1.2 TB/s Memory Bandwidth
The M5 Ultra moves 1.2 TB/s of bandwidth — enough to feed the tiny per-token slice and keep Muse Spark Open snappy on a desktop.
Safety Post-Training
Open weights get extra safety post-training: Muse Spark Open was scored against the Meta Superintelligence Scaling Framework and cleared its release bar across measured risk categories.
Muse Spark Open — FAQ
Common questions about running, licensing and downloading Muse Spark Open.
What hardware do I need to run Muse Spark Open locally?
A pair of Apple M5 Ultra Mac Studios with the newest MLX build — each supports up to 512 GB of unified memory, and together they give about a terabyte for the full model.
How large is the model and how much context does it handle?
Muse Spark Open carries 816B total parameters with 42B active, a 1M-token context window and native multimodality, aimed at coding, agentic work and long-horizon tasks.
Where can I download the weights or access the model via API?
Weights are downloadable on Amazon SageMaker AI, Google Cloud Model Garden, Hugging Face and Microsoft Foundry Models; hosted APIs run at Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter and Together AI.
Which inference engines and chips are supported?
Take your pick of MLX, llama.cpp, Unsloth, SGLang, vLLM, TensorRT-LLM or Transformers — tuned for NVIDIA, AMD, Intel and Apple hardware, with PyTorch underneath when you want to go deeper.
What license applies to the released weights?
Licensing follows the Muse Model License; its limited commercial terms still leave plenty of room for a broad range of developers to run Muse Spark Open at no cost.
How was the model evaluated for safety before release?
It was scored against the Meta Superintelligence Scaling Framework covering the most severe risks, got extra safety post-training for open weights, and cleared the release thresholds — details sit in the model card.
Get the Muse Spark Open Weights Today
Download Muse Spark Open from Hugging Face, call it through hosted inference partners, and share what you build on the Hugging Face Forums.