Meta Releases Muse Ember and Says Crouton Alpha Was an Early Version

1. 15:28
   Simon Willison

   Simon Willison Compares Meta’s Muse Ember Pricing and Reasoning Levels

   Simon Willison compares Meta’s Muse Ember 1.0 with GPT-6 Luna and Claude Haiku 5.5, showing how their input and output prices rise at different prompt length thresholds. He also reports that Ember sits on Artificial Analysis’s intelligence versus cost per task frontier. In tests generating pelican riding bicycle SVGs, the high reasoning setting failed to return a full SVG, and the other outputs did not show a clear relationship between reasoning level and visual quality.

2. 14:52
   Alexandr Wang · X

   Alexandr Wang reacts to a post about Muse Ember and Grok 4.7 pricing

   @bruno_munhoz says Muse Ember costs $0.14 and Grok 4.7 costs $8. Alexandr Wang reacts: “frame mogging” twitter.com/bruno_munhoz/stat…

3. 11:46
   Alexandr Wang · X

   Alexandr Wang Says Crouton Alpha Was an Early Version of Muse Ember

   Surprise! Crouton Alpha 🥐 was an early version of Muse Ember 🔥 We fixed a bunch of issues before launch. Thanks for all the testing! We processed over 5T tokens in testing. Excited to see where you all take this model next! twitter.com/modelsonor/status…

4. 05:58
   AI at Meta · XPicks

   AI at Meta Brings Muse Ember, Muse Wonder, and Muse Spark 1.3 to Google Cloud

   AI at Meta says it is bringing Muse Ember, Muse Wonder, and Muse Spark 1.3 to Google Cloud. The post links to an announcement about Muse Ember, but its quoted details are attributed to another post.

5. 03:56
   Alexandr Wang · X

   Alexandr Wang Encourages Users to Try Muse Ember on OpenRouter

   Give Muse Ember a go on OpenRouter 🔥

6. 03:56
   Alexandr Wang · X

   Alexandr Wang Recommends Trying Muse Ember on OpenCode

   Alexandr Wang writes, “give muse ember a go on opencode.” The quoted post from OpenCode says Muse Ember from Meta is now available in OpenCode.

7. 03:41
   The DecoderPicks

   Meta Launches Muse Ember, a Small Model for Routine Agent Tasks

   Meta has introduced Muse Ember, its smallest and most efficient AI model to date, for routine agent tasks such as tool calls, small coding tasks, request routing, data extraction, and long document processing. It handles text, images, audio, and video, with a context window of up to one million tokens. Meta says it is available through the Meta Model API and Muse Code, priced at $0.10 per million input tokens and $0.50 per million output tokens.

8. 03:41
   Alexandr Wang · X

   Alexandr Wang Says Muse Ember Is Fast, Smart, and Cost Effective

   Muse Ember is super fast, quite smart, and very cost effective. Try it out! twitter.com/rynorhn/status/21… @rynorhn says Muse Ember matches Haiku 5.5 on overall intelligence at roughly a third of the cost per task, and costs the same per token as GPT-6 Luna while scoring significantly higher. Alexandr Wang calls it super fast, quite smart, and very cost effective.

9. 03:04
   BaoyuPicks

   Meta Releases Muse Ember, Its Smallest Model, With Benchmark Cost Comparisons

   Meta released Muse Ember on November 9 as its smallest and cheapest model, aimed at repetitive agent tasks such as tool calls and small coding jobs. Developers can access it through Meta Model API or Muse Code; pricing is $0.10 per million input tokens and $0.50 per million output tokens. Artificial Analysis scores it 36 on its intelligence index and reports an average cost of $0.14 per benchmark question, compared with $0.41 for Claude Haiku 5.5; coding results vary across benchmarks.

10. 02:24
    Alexandr Wang · X

    Alexandr Wang Says Muse Ember Matches Haiku at Lower Cost and Outperforms Luna

    Alexandr Wang says Muse Ember is now on the efficient frontier 🤗 He says it matches Haiku at 3x cheaper cost per task, and is smarter than Luna.

11. 02:21
    Alexandr Wang · XPicks

    Alexandr Wang Announces Muse Ember, the Smallest Model Yet

    Alexandr Wang says Muse Ember is the smallest model they have released. He says it offers nearly Muse Spark 1.3 (max) intelligence at a 17x cheaper price point per AA, and that it is the first model they have released trained using Muse Wonder Full. The post says it is available to try on API and Muse Code.

12. 02:07
    Artificial Analysis · X

    Artificial Analysis Compares Meta’s Muse Ember on Cost and Benchmarks

    Artificial Analysis reports that Meta’s Muse Ember (max) scores 36 on its Intelligence Index at $0.14 per task, matching Claude Haiku 5.5 (max) at $0.41. Muse Ember scores 68.1% on AutomationBench-AA and performs unevenly in agentic coding, beating Claude Haiku 5.5 on Terminal-Bench 4.0 but trailing it on FrontierCode 2.0. The post lists availability through Meta’s first-party API and Muse Code.

13. 02:04
    AI at Meta · XPicks

    Meta Introduces Muse Ember for Reasoning and Agent Work

    Meta is introducing Muse Ember, which it describes as its smallest and most efficient model, built for complex reasoning execution and everyday agent work. The model can carry workflows across apps, call tools in sequence, run tests, and make targeted code fixes. It has a 1M-token context window and native understanding of text, images, audio, and video.

14. 02:00
    TechCrunch · AI

    Meta Releases Muse Ember, a Lightweight Model for Agent Execution

    Meta Superintelligence Lab released Muse Ember, which it describes as its lightest and most cost-efficient AI model yet, for agent tasks such as tool calls, coding subtasks, routing, extraction, and long document parsing. Meta says Ember costs 10 cents per million input tokens and is designed to pair with a larger model such as Muse Wonder. The company also says it handles text, images, and audio.

Trends compare the same continuously observed participants, which may cover fewer sources than current heat. Point or click to inspect hourly heat, or use left and right arrow keys.