The models add “active memory” and “active focus” capabilities.
After spending more than a year and tens of billions of dollars building up its AI efforts, Meta says it has caught up. The company introduced its two latest AI models, Muse Wonder and Muse Wonder Full, both of which represent “a giant leap towards personal superintelligence”.
According to Meta, the models are particularly adept at navigating legal, financial and health-related tasks, and are capable of “delivering borderline coding performance.” Both models also add “active memory” and “active focus,” which Meta says help people accomplish more over long periods of time.
That’s notable because Meta says it plans to bring Wonder to its personal agent Muse, starting with paid subscribers. Active memory, according to the company, is “the ability to carry what’s important to you over conversations and recall details relevant to the task at hand” while active focus allows an agent to “keep working on your goal when a task becomes difficult.”
The models are a notable milestone for Meta, which has said it expects to spend as much as $145 billion this year to fuel its ambitious AI plans. In an interview with the Financial TimesCEO Mark Zuckerberg said that “getting to this point required both a large amount of computing and investment, and a lot of very rigorous scientific work” by the company’s superintelligence team.
The new models are now being released to subscribers. Muse Wonder is available via Meta’s API and in Muse Code. It is priced at $4 per million input tokens and $20 per million output tokens. Muse Wonder Full is available to people who subscribe to the $50/month “power usage” or $200/month “maximum usage” tiers of Muse Code.