Meta has introduced Muse Ember via its Superintelligence Lab, a lightweight multi-modal model built for complex reasoning, tool calls, and long-document parsing across text, images, and audio at 10 cents per million input tokens.
The new system was designed for complex thinking, use of tools, programming small pieces of code, managing requests, extracting information, and analyzing large documents.
Low cost and collaboration with larger systems
Meta reported that the cost is ten minutes per million data points in input. According to comparisons using common evaluation criteria, Ember is cheaper and more capable than Anthropic's Haiku and OpenAI's Luna, while being in the same price range as open models such as DeepSeek V4.1 Flash and GLM 5.3 Flash.
The model is intended to work in conjunction with larger Meta systems, such as Muse Wonder. For example, in customer service, a larger system could define the rules and Ember could apply them to many different cases.
Support for different media formats
Ember was trained to handle text, images, and audio. This allows it to process screenshots, diagrams, scanned documents, voice messages, meeting recordings, and videos without the need for a separate add-on system.
↳
READ ALSO
HACKS.GR / RELATED
AI Unlocked
Nvidia drops 8% and loses 350 billion after investigation into Jensen Huang
↗
Ember is the fourth model in the Muse series, following Spark in April, Glimmer in August and Wonder in October.
>Stay connected with Hacks.gr
Set us as a preferred source on Google and find us on social media.
Preferred source
↗
end of article
Top comments (0)