Meta launches Muse Ember, its most efficient AI model to handle repetitive tasks; know what it can do and here's how it works

Meta has said that Muse Ember is significantly cheaper than competitors like Haiku and Luna. Despite being designed as a lightweight execution model, Ember can process different types of information. Meta claims it supports text, images and audio....

AP
Meta launches Muse Ember.
Meta Superintelligence Lab on Monday released a new AI model called Muse Ember. It is the lab's smallest and most efficient AI model to date. Meta said that it was "designed for complex reasoning execution and the work agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing." Meta has also claimed that Muse Ember delivers more intelligence per dollar at every reasoning level than Anthropic's Haiku and OpenAI's Luna. The objective is to make AI agents faster and cheaper to operate.

What makes Muse Ember different from other AI models?

Meta CEO Mark Zuckerberg had earlier said that it is "Our smallest model that delivers higher intelligence per dollar than others in its class. Built for token-heavy work that agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing. Available today on the Meta Model API and in Muse Code."

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Releasing Muse Ember today -- our smallest model that delivers higher intelligence per dollar than others in its class. Built for token-heavy work that agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing. Available today on the Meta Model API and in Muse Code.</p>— Mark Zuckerberg (@finkd) <a href="https://x.com/finkd/status/2119851180948152590?ref_src=twsrc%5Etfw">November 9, 2026</a></blockquote><script async="" src="https://platform.x.com/widgets.js" charset="utf-8"></script>
Muse Ember is significantly cheaper than competitors like Haiku and Luna. This is what makes it one of the best selling points. It costs 10 cents per million input tokens. According to TechCrunch, Meta considers Ember cost-competitive with open-source alternatives such as DeepSeek V4.1 Flash and GLM 5.3 Flash.


It is designed to work alongside its more powerful Muse Wonder model. Meta claims that "Muse Wonder plans, Muse Ember handles the volume."

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">1/ announcing our smallest, most efficient little model is here: muse ember ��built for work that agents repeat all day, it gets you more intelligence per dollar than Haiku, Luna, and all other flash models at any reasoning setting and does it at a third of the cost or less! https://t.co/ClmgPOTrsR</p>— Alexandr Wang (@alexandr_wang) <a href="https://x.com/alexandr_wang/status/2119853374158032950?ref_src=twsrc%5Etfw">November 9, 2026</a></blockquote><script async="" src="https://platform.x.com/widgets.js" charset="utf-8"></script>

What are Muse Ember's capabilities?

Despite being designed as a lightweight execution model, Ember can process different types of information. Meta claims it supports text, images and audio. Meta had put it in the announcement that "agents can read screenshots, charts, and scanned documents, or work from voice memos, meeting recordings, and video without a separate model."

Muse Ember is the fourth model released under Meta's Muse brand. It follows Muse Spark, released in April 2026; Muse Gimmer, released in August; and Muse Wonder, released in October. Meanwhile, Muse Agent was also launched in September. It had new features like Spaces and video chat rolling out earlier this month.
Download
The Economic Times Business News App
for the Latest News in Business, Sensex, Stock Market Updates & More.
Download
The Economic Times News App
for Quarterly Results, Latest News in ITR, Business, Share Market, Live Sensex News & More.
READ MORE
ADVERTISEMENT

READ MORE:

LOGIN & CLAIM

50 TIMESPOINTS

More from our Partners

Loading next story
Business News › News › International › US News › Meta launches Muse Ember, its most efficient AI model to handle repetitive tasks; know what it can do and here's how it works
Text Size:AAA
Success
This article has been saved

*

+