What Meta released
Meta's research blog, dated November 9, 2026, states: "Today, we're introducing Muse Ember, our smallest and most efficient model designed for complex reasoning execution and the work agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing." It is available "on Muse Code and Meta Model API as muse-ember-1.0."
Pricing in the availability section is "$0.10 per million input tokens, $0.004 per million cached input tokens and $0.50 per million output tokens, with a 1M-token context window and up to 512K output tokens." A footnote says that pricing is for prompts up to 200K tokens; longer prompts cost $0.30 input, $0.012 cached input and $1.50 output per million tokens. The model supports the Responses API and, Meta writes, "natively understands text, images, audio, and video."
How Meta positions it
Meta says Muse Ember "is the first trained using Muse Wonder Full" and places it beside Muse Wonder for harder problems and Muse Spark for coding and agentic workflows. Ember is meant for repeatable tasks at scale. The post also says Muse Ember, Muse Spark and Muse Wonder are now available on Google Cloud and Oracle Cloud.
What Meta reports
Meta writes that "Muse Ember delivers more intelligence per dollar than Claude Haiku 5.5 or GPT-6 Luna at any reasoning setting: it matches Haiku's best score at about a third of the cost and beats every Luna setting." An accompanying chart is labeled as Artificial Analysis results, with a note that "Muse Ember costs are estimates at proposed pricing" in one caption and "cost at list pricing" in another. In Meta's results table, the Artificial Analysis Index at max is 36 for Muse Ember and Claude Haiku 5.5, and 28 for GPT-6 Luna. On FrontierCode 2.0-AA, Haiku 5.5 is listed higher (34.9 against Ember's 27.2). The Personal Superintelligence Index is described as Meta's internal composite. These figures have not been independently verified here. Partner quotes in the post are customer opinions, not measured results.
What it means for you
If you run many short agent steps or subagents, Meta is offering a cheaper, multimodal API model to try against Haiku 5.5 or GPT-6 Luna on your own tasks. Check the 200K prompt price step before sending very long contexts. AppFor.ai has not run Muse Ember.
Go to the source
Sources & further reading
Based on the official documentation below. Features and access can change; check the provider before making a decision.
Was this page useful?
Help us make our coverage more useful. This is feedback on the page, not a rating of the tool.
Go a little further