Meta debuts pair of frontier models, Muse Wonder and Muse Wonder Full

Meta's Superintelligence Lab on Oct. 12 released Muse Wonder and Muse Wonder Full, a pair of new state-of-the-art models and the company's first attempt at a frontier model. Unlike the lightweight Muse Spark, the Wonder models are larger and designed for higher benchmark scores and better performance on complex tasks such as coding.

Meta said its internal testing shows the two models achieve benchmark scores comparable to Fable 5.1 and GPT-6 Astra. The company described the release as "the next step on our scaling ladder and a major leap toward personal superintelligence."

Meta placed particular emphasis on the models' ability to retain and synthesize user-provided information over time, a capability the company calls "active memory." According to Meta, Muse Wonder scores higher than Fable, Opus and Astra on the MemoryBench benchmark.

Wonder is available through the Meta Model API at $4 per million input tokens and $20 per million output tokens, roughly equivalent in cost to GPT-6 Sol or Opus 5.5. Wonder Full will only be offered to Muse Code subscribers and will not be served through the API. Both models will also power the Muse agent.

The models were trained at Meta's Prometheus data center in central Ohio, which came online earlier in 2026. The release is Meta's first frontier model launch since Alexandr Wang restructured the company's AI efforts in August 2025.