AI Model Release Tracker

Muse Ember

Released

Muse Ember is an AI model released by Meta on Monday, Nov 9 2026, 25 days after Muse Spark Open. It has a 1M token context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider.. Benchmark results (shown below) cover AutomationBench (Artificial Analysis), GDP.PDF (Artificial Analysis), AA Coding Agent Index, Terminal-Bench 4.0, and AA Intelligence Index v5.0.

API pricing

Up to 200K input tokens
InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it.
$0.10
Cached inputA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again.
$0.004
OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one.
$0.50
Over 200K input tokens
InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it.
$0.30
Cached inputA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again.
$0.012
OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one.
$1.50
  • Reasoning effort runs from "minimal" to "max". When a request sets none, the default is "max".
  • A request with more than 200K input tokens, cached tokens included, is billed entirely at the long-context rates.
  • Standard mode only on the Model API. Contributor mode is not available for Muse Ember.
USD per 1M tokensEvery price here is for one million tokens. A token is roughly three-quarters of a word, so a million tokens is about 750,000 words of text.Model mappingVerified November 9, 2026Read from dev.meta.ai. Click the check to open the source.

Benchmarks

Coding

AA Coding Agent IndexAgentic coding — Artificial Analysis' overall score for coding agents, combining three coding benchmarks with what each run costs and how many tokens it burns. It rates a model paired with a particular agent harness rather than the model alone, so the same model scores differently in different tools. Higher is better.
46
#3 of 3

Terminal & CLI

Terminal-Bench 4.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better.
40.9%
#19 of 36

Agentic & tool use

AutomationBench (Artificial Analysis)Business workflows — The same business-workflow test, run by Artificial Analysis rather than by the lab, and graded strict pass@1: a workflow counts only if the whole of it is right on the first attempt. Read it as a separate board from the lab-run AutomationBench scores. Higher is better.
37.8%

Knowledge work

AA Intelligence Index v5.0Overall intelligence — Artificial Analysis composite intelligence index, version 5.0 with tools; separate from earlier index editions
36
#2 of 2

Multimodal

GDP.PDF (Artificial Analysis)Document comprehension — The same professional-PDF test, run by Artificial Analysis rather than by the lab. Read it as a separate board from the lab-run GDP.PDF scores. Higher is better.
22.8%

Compare Muse Ember with

Muse Ember

Suggested comparisons

Muse EmbervsMuse Spark OpenMuse EmbervsGPT-6.1 TerraMuse EmbervsClaude Fable 5.2Muse EmbervsGemini 4 CarbonMuse EmbervsGrok 4.7Muse EmbervsDeepSeek-V4.1-Flash-1017Muse EmbervsMistral Large 4Muse EmbervsKimi K3.1 ProMuse EmbervsGLM-5.3-FlashMuse EmbervsQwen3.8-Max-0902Muse EmbervsNemotron 3.5 Bolt E4B 4GB

About

Muse Ember was introduced on November 9, 2026 as Meta's smallest Muse model, built for the work agents repeat: tool calls, coding subtasks, routing, extraction and long-document parsing. It was available the same day on the Meta Model API, under the id muse-ember-1.0, and in Muse Code on pay-as-you-go and every subscription plan. Meta called it the first model trained using Muse Wonder Full. The docs give it a 1,048,576-token context window, up to 524,288 output tokens, and text, image, video, audio and PDF inputs, and they pair it with Wonder or Spark when a workflow needs a larger model to plan.

Launch numbers are for max reasoning, which is the default effort when a request sets none. Meta reported 40.9% on Terminal-Bench 4.0 using mini-swe-agent, the harness it called Ember's best published one, 46 on the Artificial Analysis Coding Agent Index and 36 on Intelligence Index v5. Artificial Analysis runs also gave it 37.8% on AutomationBench and 22.8% on GDP.pdf. Standard token rates are $0.10 input and $0.50 output per million, with cached input at $0.004, for requests of up to 200,000 input tokens; above that, the whole request is billed at $0.30 input, $0.012 cached and $1.50 output. No parameter count or weight download was published, and the Model API offers Ember in Standard mode only.

Frequently asked questions

Muse Ember was released by Meta on Monday, Nov 9 2026.

All Meta releases

19 tracked