AI BriefingMeta launches Muse Wonder model series for personal and professional agents
Key point
The new Muse Wonder series introduces Active Memory and Active Focus features, available now via the Meta Model API and Muse Code.
Details
Meta has launched Muse Wonder, a new model series designed for personal and professional agents. The release includes two variants: Muse Wonder and Muse Wonder Full, which are described as Meta's most capable models for agent-based tasks.
Key Features
The series is built around four core capabilities aimed at improving long-running, personal work:
- Active Memory: Carries context across conversations and retrieves relevant details for current tasks.
- Active Focus: Maintains progress on hard goals by working through blockers and asking for clarification before important actions.
- Domain Expertise: Specialized knowledge in law, finance, and health (e.g., understanding leases).
- Frontier Coding: Tracks requirements, edits codebases, and iterates through testing cycles.
Availability and Performance
Muse Wonder is available immediately through the Meta Model API and Muse Code. Muse Wonder Full is exclusive to Muse Code subscribers. Meta claims state-of-the-art results on benchmarks including Harvey’s Legal Agent Benchmark, FocusBench, and SWE-Marathon, though specific scores are not provided in the announcement text.
Safety
Meta states the models include its strongest safeguards to date against chemical, biological, radiological, nuclear, and cyber misuse, operating under its Superintelligence Scaling Framework.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.