Meta's Superintelligence Lab has launched Muse Wonder and Muse Wonder Full, larger and more capable models than the earlier Muse Spark. The company calls them its first attempt at a frontier model.
Meta released the two models on Monday. Unlike Muse Spark, which was built as a lightweight, low-latency model, the Wonder models are larger and are said to score higher on benchmarks and handle complex tasks such as coding better. Meta's own testing, presented in the model card, shows results comparable to Fable 5.1 and GPT-6 Astra.
Muse Wonder is available through the Meta Model API at $4 per million input tokens and $20 per million output tokens, which puts it roughly on par with GPT-6 Sol or Opus 5.5. Wonder Full is different: it will be offered only to Muse Code subscribers and will not be served through the API. Both models will also power the Muse agent.
Meta put particular emphasis on what it calls "active memory," the ability to retain and synthesize information a user provides. The company points to the MemoryBench benchmark, where it says Muse Wonder outscores Fable, Opus and Astra. Research scientist Kavya Kopparapu said in a launch video that the team worked hard to make sure the model can remember what matters to users over days, weeks and months.
The release is the first frontier model since Alexandr Wang, founder of Scale AI, restructured Meta's AI efforts in August of last year. Since then, Meta had shipped Muse Spark and the Muse agent but had not put out the kind of state-of-the-art models associated with OpenAI and Anthropic. According to Meta, the new models were trained at its Prometheus data center in central Ohio, which came online earlier this year.
Source: TechCrunch
