Meta launches Muse Wonder models targeting personal and professional AI agents
CurieuxExplorer · x · 2026-10-13
- Meta officially introduced Muse Wonder and Muse Wonder Full, frontier models claiming state-of-the-art performance on personal and professional agentic tasks — billed as the next step toward personal superintelligence.
- Key technologies include Active Memory (tracking what matters to the user) and Active Focus (diligently working toward user goals); the research page also covers coding capabilities, benchmarks, and demos.
- Context: Meta says the release follows a ground-up overhaul of its AI stack over the past 14 months, building on the scaling approach laid out with Muse Spark across pretraining and reinforcement learning.
- Availability: Muse Wonder is live via the Meta Model API and Muse Code; Muse Wonder Full is for Muse Code subscribers. It will roll out to Muse users first, powering artifact generation and other high-intelligence workflows, and will seed future specialized models.
More from coding & agent
- RTNav: open-source modular asynchronous agent for real-time zero-shot object navigation — rsasaki0109 · 2026-10-13
- Claude Code token-saving tips: 5 settings that keep cache hit rate above 98% — AlchainHust · 2026-10-13
- This developer built an Obsidian Agent that cleans the inbox with one sentence — 17 files sorted in 2 minutes — vista8 · 2026-10-13
- LLM-free Jev pipeline diagnoses K8s faults in 14.6s, passing 76.2% of 105 diagnoses — bibryam · 2026-10-13
- Agents get their own Mac mini, email and Telegram: one user's isolation boundary setup — 1FunkyFriedChicken · 2026-10-13
- jitprompt: An MCP server that serves agent instructions just-in-time instead of one giant system prompt — mirkop82_ · 2026-10-13