Meta rolls out Muse Spark, its latest AI model for everyday tasks

Meta has unveiled Muse Spark, its latest artificial intelligence model aimed at everyday consumer use, marking the first major product from its Superintelligence Labs initiative.

Chief executive Mark Zuckerberg announced the launch in a public post, saying the model now powers an updated version of Meta AI, available through the company’s app and web interface, as reported by Mashable.

“Muse Spark is the first step on our scaling ladder and the first product of a ground-up overhaul of our AI efforts,” Meta said in its announcement.

Focus on everyday use cases

Spark is designed to handle practical, day-to-day tasks, including visual understanding, health-related queries, shopping assistance and social content interactions. According to Mashable, the model is positioned as a step towards more advanced systems that move beyond answering questions to acting as digital agents.

Zuckerberg said future iterations would focus on AI systems that “do things for you”, signalling a broader shift towards automation-led personal assistants.

Meta also indicated that additional models in the Muse family will include open-source releases.

First output from a broader AI overhaul

The launch represents the first tangible outcome of Meta’s wider push into advanced AI, following the formation of its Superintelligence Labs in 2025.

As reported by Mashable, the initiative has undergone rapid expansion and restructuring over the past year. Meta recruited more than 50 researchers from competitors including OpenAI, Anthropic and Google, and appointed former Scale AI chief Alexandr Wang to lead the effort.

The company later paused hiring and reorganised teams into smaller units focused on research, product development, infrastructure and superintelligence.

Zuckerberg has previously argued that compact teams are better suited to delivering breakthroughs than larger, more fragmented organisations.

Investment scale and market pressure

The rollout comes amid significant financial commitments to AI. Meta allocated around $72 billion to AI development in 2025 and is expected to increase spending to as much as $135 billion in 2026, according to Mashable.

The scale of investment has drawn scrutiny, particularly as questions persist over the commercial returns of AI deployments. An MIT study cited by Mashable found that most companies using AI have yet to see meaningful financial gains.

Against this backdrop, Muse Spark is being closely watched as an early test of Meta’s ability to convert spending into viable products.

Performance and competition

Meta released benchmark data comparing Spark with leading models, including those from OpenAI, Google and Anthropic.The results were mixed, with the model outperforming rivals in some tests while trailing in others.

The company also introduced a “Contemplating” mode designed to improve reasoning by coordinating multiple AI agents, although this feature is yet to be widely available.

The Muse programme reflects Zuckerberg’s longer-term vision of “personal superintelligence”, an approach that emphasises individual empowerment rather than centralised AI systems.

While the company’s earlier Llama models have lagged behind competitors on several benchmarks, the launch of Spark signals a shift from experimentation to product delivery.

Meta said it would roll out additional features and models over time. The market response will likely hinge on whether Spark can match the capabilities of rival systems and demonstrate clear user value as competition in AI intensifies.