Meta rolls out specialized models for Muse and Meta AI
Meta has begun rolling out specialized models trained with Muse Wonder to Muse and Meta AI; the company says the current Muse model reaches 47% on its personal-agent index.
AT A GLANCE
Key points
- Meta is rolling specialized models trained with Muse Wonder into Muse and Meta AI.
- Meta reports a 47% Personal Superintelligence Index score, up from 38% for the previous Muse model.
- The company reports 90% on its Active Memory evaluation and a 67% improvement on obstacle-heavy everyday tasks.
What is changing?
Meta announced that it has begun rolling out specialized models trained with capabilities learned from Muse Wonder to people using Muse and Meta AI. On the company’s own Personal Superintelligence Index, the current Muse model scored 47%, compared with 38% for the previous model.
The update brings the Active Memory and Active Focus capabilities introduced with Muse Wonder into broader product experiences. Meta says the current model reached 90% on tasks that require critical context, versus 76% for the previous model. The company also reports a 67% improvement in completing everyday tasks that contain unexpected obstacles.
Active Memory is intended to separate temporary from lasting information and retrieve still-relevant context from earlier conversations. Active Focus is designed to try alternative paths when obstacles interrupt progress toward a goal. In Meta AI, the same model family will be used to break complex work into steps and adapt when the user changes direction.
Explore the original sources
Source date: 6 November 2026.
Published by AI Radar: 7 Nov 2026, 06:01 (Istanbul, UTC+3).
FOLLOW THE TOPIC
Topics and tags
Share your discovery.
Send this story to a friend.
Reader comments
What do you think about this development? Share your experience or question. Comments are published after editorial approval.
Your name/nickname and comment become public if approved. Read: Privacy notice.
Loading comments…