Meta‘s Superintelligence Lab released two new state-of-the-art models on Monday: Muse Wonder and Muse Wonder Full, calling them the company’s first attempt at a frontier model. The earlier Muse Spark was positioned as a lightweight, low-latency model, while the Wonder models are larger and, according to Meta, more capable, with higher benchmark scores and better results on complex tasks such as coding. In testing reported in the model card, Meta says Wonder and Wonder Full achieve benchmark scores comparable to Fable 5.1 and GPT-6 Astra. Meta‘s blog post describes the release as “the next step on our scaling ladder and a major leap toward personal superintelligence.”

The models’ scale is reflected in their pricing. Accessed through the Meta Model API, Wonder will cost $4.00 per million input tokens and $20.00 per million output tokens, making it roughly equivalent in cost to GPT-6 Sol or Opus 5.5. The larger Wonder Full model will not be served through the API; it will be available only to Muse Code subscribers. Both models will also be used to power the Muse agent.

Meta put particular emphasis on the new models’ “active memory,” a term the company uses for the ability to retain and synthesize information provided by the user over long periods. To support that goal, Meta highlighted the MemoryBench benchmark, which evaluates continual learning and memory construction frameworks. According to Meta‘s testing, Muse Wonder scores higher on MemoryBench than Fable, Opus, or Astra. Research scientist Kavya Kopparapu said in a launch video: “We put a tremendous amount of effort into making sure your model can remember what’s important to you over days, weeks, months and beyond.”

The Wonder release is notable as Meta‘s first frontier model release since Scale AI founder Alexandr Wang restructured the company’s AI efforts in August of last year. Since then, Meta had shipped the lightweight Muse Spark and the Muse agent but avoided the large state-of-the-art model launches associated with OpenAI and Anthropic. According to the post, the models were trained at Meta‘s Prometheus data center in central Ohio, which came online earlier this year.