Meta has introduced the Muse Wonder series, featuring Muse Wonder and Muse Wonder Full, which the company describes as its most capable models for personal and professional agentic tasks. Muse Wonder is available today through the Meta Model API and Muse Code, while Muse Wonder Full is available to Muse Code subscribers. The models mark the next step on Meta's scaling ladder and a major leap toward personal superintelligence.
The series combines two core capabilities, Active Memory and Active Focus. Active Memory lets the model carry important context about users across conversations and retrieve relevant details when needed. Active Focus describes how the model keeps working toward difficult goals by trying alternative approaches, working through blockers, adapting to new information, asking for clarification and checking in before important actions.
Meta says the models achieve state-of-the-art results across key benchmarks for personal and professional agents, memory and coding. Evaluations cover domain expertise in law, finance and health, including Harvey's Legal Agent Benchmark, which tests whether models can work from source documents to produce legal work products. Focus performance is measured through Focus Bench and SWE-Marathon, a test of long-running software projects.
Muse Wonder was trained at Meta's gigawatt-scale Prometheus data center and builds on the company's earlier Muse Spark release. Muse Wonder will soon power experiences in Muse, rolling out first to subscribers, and will also drive artifact generation and more intensive processes. Meta plans to use Muse Wonder to develop specialized models for broader use across Muse.
Source: Hacker News · Summarized by HeadlinesBriefing