Skip to content

Related video: MIT researcher shows how robots learn physical tasks through simulation and practice, Imagination in Action

Trending on TechCrunch AI

Meta launches Muse Ember, lightest and most cost-efficient model yet

Techmeme · · 20-second read

Share

    November 9, 2026 · AI Tech

    Meta's Superintelligence Lab released Muse Ember, a new AI model designed to be the company's smallest and most cost-efficient to date. The model is positioned as delivering more intelligence per dollar than competing offerings from other major AI labs.

    The release represents Meta's continued investment in a range of model sizes across its AI portfolio, from large frontier models down to compact, efficient variants suited for resource-constrained deployments.

    Read the full story at Techmeme

    This is an AI-written summary. The original reporting belongs to Techmeme. How we write these · Spot a mistake?

    Get the AI brief every morning.

    The five AI stories that matter each morning: startups, money, jobs, tech and what's coming next. Free, and one click to unsubscribe. See our Privacy Policy.