Meta Enters the Frontier Model Race
Meta's Superintelligence Lab released a pair of new state-of-the-art models on Monday โ Muse Wonder and Muse Wonder Full โ marking the company's first genuine attempt at a frontier model.
Where the earlier Muse Spark was designed as a lightweight model optimized for low-latency performance, the Wonder models are substantially larger and more capable. They post higher benchmark scores and deliver better results on complex tasks such as coding.
Competitive Benchmarks
According to Meta's internal testing, detailed in the model card, both Wonder and Wonder Full achieve benchmark scores comparable to Fable 5.1 and GPT-6 Astra.
"Together, they mark the next step on our scaling ladder and a major leap toward personal superintelligence," Meta wrote in its blog post announcing the models.
That scale is reflected in the pricing. Accessed through the Meta Model API, Wonder costs $4.00 per million input tokens and $20.00 per million output tokens โ roughly equivalent to GPT-6 Sol or Opus 5.5.
Notably, Muse Wonder Full will only be available to Muse Code subscribers and will not be served through the API. Both models will also power the Muse agent.
A Focus on "Active Memory"
In keeping with Meta's ambition to build personal assistants capable of performing tasks over long periods, the company placed particular emphasis on the new models' ability to retain and synthesize user-provided information โ a property Meta calls "active memory."
To demonstrate improved recall, Meta highlighted the MemoryBench benchmark, which challenges continual learning and memory construction frameworks. According to the company's testing, Muse Wonder now scores higher on MemoryBench than Fable, Opus, or Astra.
"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," said research scientist Kavya Kopparapu in a video accompanying the launch.
The First Release Since Meta's AI Restructuring
The new models represent the first frontier model release since Scale AI founder Alexandr Wang restructured Meta's AI efforts in August of last year. Since then, Meta has shipped its lightweight Muse Spark model and the Muse agent, but had avoided the kind of state-of-the-art releases OpenAI and Anthropic are known for.
According to the blog post, the models were trained at Meta's Prometheus data center in central Ohio, which came online earlier this year.
via TechCrunch AI