Meta has spent more than a year rebuilding its AI operation around one ambitious phrase: personal superintelligence. Now it has a model that will show whether the rebuild actually worked.
On October 12, 2026, the company introduced the Muse Wonder series through its AI Research blog. The release includes two models, Muse Wonder and Muse Wonder Full. Bloomberg framed the launch as a test of the capabilities of Meta’s superintelligence team, and that framing fits.
What Muse Wonder actually does
The Muse Wonder models are built for agentic tasks, in both personal and professional settings.
Two features carry most of that load. Meta calls them Active Memory and Active Focus, and they are designed to handle complex workflows that stretch over long periods.
The models have been benchmarked as state-of-the-art in three areas: memory, coding, and agent performance.
Availability is immediate. Both models can be accessed through the Meta Model API and Muse Code, with the Full version reserved for subscribers.
Training took place at Prometheus, Meta’s gigawatt-scale data center.
The team behind the model
Mark Zuckerberg announced Meta Superintelligence Labs, or MSL, on June 30, 2025. The restructuring began around mid-2025 and pulled several existing teams under one roof, including FAIR, Meta’s long-running and well-known research lab.
Leadership came from outside the company. Alexandr Wang, previously of Scale AI, serves as Chief AI Officer. Nat Friedman, formerly of GitHub, is the product lead.
By late 2025, MSL had somewhere around 3,000 to 3,400 staff, including high-profile hires from rivals such as OpenAI and Google DeepMind.
A step-by-step rollout
The series succeeds Muse Spark, which debuted in April 2026. In July 2026, Meta added Muse Image and Muse Video to the family. Wonder now sits on top of that stack as the more advanced, agent-focused option.
What this means for Meta and the AI race
Distribution is Meta’s quiet advantage. The company describes the models as aimed at personal superintelligence, with the potential to serve billions of users.
Meta has historically made money from advertising rather than charging users directly. A paid AI tier for Muse Wonder Full would mark a meaningful addition to how the company monetizes its products.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.