Jakarta, INTI - Meta has officially launched its latest artificial intelligence (AI) model, Muse Spark, marking a serious push to compete in a market currently dominated by players like OpenAI and Google through products such as ChatGPT and Gemini.
Muse Spark is the first model in the Muse series developed by Meta Superintelligence Labs, the company’s AI division led by Alexandr Wang, who joined Meta following a major investment of US$14.3 billion in Scale AI.
Meta revealed that it has overhauled its AI development strategy over the past nine months, following lukewarm reception to its earlier open-source models. CEO Mark Zuckerberg has since shifted focus toward building more efficient AI systems capable of handling complex tasks across domains such as science, mathematics, and healthcare.
According to Meta, its Superintelligence Labs rebuilt the entire AI stack from the ground up, accelerating development at an unprecedented pace. Rather than positioning Muse Spark as a flagship model, the company designed it as a lightweight yet powerful system that prioritizes speed and efficiency while maintaining competitive performance in reasoning and multimodal processing tasks.
Racing to Catch Up in the AI Market
Competition in the AI industry continues to intensify. Beyond OpenAI, major players such as Anthropic and Google have rapidly advanced, pushing the combined valuation of the AI sector beyond US$1 trillion.
The global generative AI market is also experiencing rapid growth, projected to expand by more than 40 percent annually. It is expected to grow from approximately US$22 billion in 2025 to nearly US$325 billion by 2033.
To close the gap, Meta is significantly ramping up its capital expenditure. In 2026, the company plans to invest between US$115 billion and US$135 billion in AI infrastructure, almost double its spending from the previous year.
Meta claims that Muse Spark can deliver performance comparable to earlier models such as Llama 4 while requiring significantly lower computational resources. This improvement is attributed to advancements in AI training techniques and upgraded infrastructure.
The model demonstrates strong capabilities in multimodal perception, reasoning, healthcare-related tasks, and agent-based operations. However, Meta acknowledged that further improvements are still needed, particularly in long-term agent systems and coding workflows.
At the same time, Meta is beginning to explore new business opportunities around Muse Spark. The company is currently testing limited API access with selected partners, with plans to introduce paid services for developers in the future, mirroring monetization strategies already adopted by OpenAI and Google.
Integration Across Meta’s Ecosystem
Muse Spark will be deeply integrated into Meta’s ecosystem, including platforms such as Facebook, Instagram, WhatsApp, and Messenger, as well as devices like Ray-Ban Meta Smart Glasses.
The model will enhance Meta AI services through multiple modes, including a fast mode for simple queries, an analysis mode for complex tasks such as legal document review, and a reasoning mode designed for advanced multi-agent problem solving.
These capabilities are intended to rival advanced reasoning systems like Gemini Deep Think and GPT Pro.
In addition, Meta is introducing new features such as a Shopping Mode, allowing users to discover fashion inspiration, purchase items, and even design living spaces based on AI-driven recommendations.
Conclusion
Meta’s launch of Muse Spark reflects its renewed strategy to compete in the rapidly evolving AI landscape. By prioritizing efficiency, scalability, and deep integration across its ecosystem, Meta aims to close the gap with industry leaders while unlocking new opportunities in AI-driven services and monetization.
Read more: An AI Agent Hacked FreeBSD OS Without Human Help, Only Took 4 to 8 Hours