TLDR
- Meta Platforms (META) closed at $736.72 before sliding to $735.02 in extended trading.
- The new Muse Spark Open features 816 billion parameters with 1M-token context capability.
- Two M5 Ultra Mac Studio machines are required for local deployment of the AI system.
- Multiple cloud providers and API platforms now support Muse Spark Open access.
- The release follows Meta’s August introduction of the smaller Muse Glimmer model.
Meta Platforms experienced a modest decline Thursday, October 15, coinciding with the debut of Muse Spark Open, its latest artificial intelligence system tailored for programming and automation workflows. The stock (META) settled at $736.72, marking a 0.11% decrease, before dropping an additional 0.23% to $735.02 during after-hours trading. The introduction of this advanced AI tool came alongside back-to-back trading session losses for the social media giant.
Meta Unveils Muse Spark Open AI System for Software Engineering
Meta rolled out Muse Spark Open as its newest open-weight artificial intelligence framework targeting software creation and complex analytical reasoning. The architecture incorporates 816 billion total parameters while activating 42 billion during actual computation. The platform manages context windows extending to one million tokens, enabling it to tackle extended and intricate development assignments.
This AI solution processes diverse data types and executes programming operations requiring sequential decision-making. According to Meta, the technology accommodates extended project timelines and challenging engineering workflows. Internal benchmark testing positioned its capabilities against competing open-weight alternatives currently available to the developer community.
Open-weight architectures enable developers to obtain model parameters and tailor systems for specialized use cases. Organizations gain deployment autonomy and can maintain data within their chosen infrastructure. Meta’s August release of the compact Muse Glimmer preceded this more substantial offering.
Apple Mac Studio Configuration Carries Significant Hardware Requirements
Meta engineered Muse Spark Open for operation across dual Apple Mac Studio units equipped with M5 Ultra processors. The arrangement leverages Apple’s MLX framework for handling machine learning tasks on-premises. This configuration demands considerable computing power, including approximately one terabyte of system memory.
The technology company evaluated Muse Spark Open in conjunction with Muse Code, its command-line development assistant. A practical demonstration involved modifying a Mastodon application through bug corrections and feature additions. These trials showcased the system’s ability to execute interdependent operations without requiring constant human intervention.
Meta further assessed the model using DaVinci Resolve, a professional-grade video production tool. During testing, the AI analyzed video files and performed selective trimming operations. This application highlighted the platform’s versatility beyond traditional software engineering scenarios.
Cloud Platform Integration Broadens Developer Accessibility
Meta enabled access to the model via Baseten, Fireworks AI, Groq, Modal, Nebius, OpenRouter, and Together AI. Developer teams can retrieve model weights through Hugging Face, Amazon SageMaker AI, Google Cloud, and Microsoft Foundry. These distribution channels offer options beyond deploying costly on-site hardware infrastructure.
Supported inference platforms encompass vLLM, SGLang, TensorRT-LLM, llama.cpp, and Hugging Face Transformers. The system accommodates AMD, Apple, Intel, and Nvidia processing architectures for broader implementation flexibility. Teams can fine-tune the framework through PyTorch for domain-specific requirements.
Open-weight availability may decrease reliance on closed-source alternatives while enhancing deployment adaptability. Nevertheless, operating large-scale models continues to require advanced technical knowledge and substantial computational infrastructure. Meta’s announcement adds to its AI product portfolio as Thursday’s market activity reflected modest weakness.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
