Meta releases lightweight Muse Ember model for agent execution
Meta Superintelligence Lab released Muse Ember, the fourth model under its Muse brand and its lightest, most cost-efficient yet, at 10 cents per million input tokens. Aimed at high-volume agent work such as tool calls, coding subtasks, routing and document parsing, it handles text, images and audio and is designed to pair with the larger Muse Wonder.