Meta · Available now

Meta Muse Ember price per reply in 2026: what it costs and what it does

Muse Ember is Meta’s smallest and most efficient Muse model, released on November 9, 2026 with a 1M-token context window, at $0.10 per million input and $0.50 per million output tokens for prompts up to 200K tokens on Meta’s API.

Available nowLast updated:

Use it now

Chat

≈ 1 credit per reply · See all models

New accounts get 5 free credits once, valid for 7 days. They work on the lower-priced models here; the priciest ones need credits you bought.

Start chatting — 5 free credits
Maker
Meta
Type
Text
Released
2026-11-09
OpenRouter id
meta/muse-ember

Per million tokens

Price

Meta Muse Ember · checked 2026-11-17
-InputOutputSource
Meta — Prompts up to 200K tokens$0.10$0.50research.meta.ai
Meta — Prompts over 200K tokens$0.30$1.5research.meta.ai
Via OpenRouter (meta/muse-ember)$0.10$0.50openrouter.ai

Checked

Facts, with sources

  • Meta’s post "Introducing Our Smallest, Most Efficient Model: Muse Ember" is dated November 9, 2026 and says Muse Ember is "available today" on the Meta Model API and in Muse Code, as muse-ember-1.0. It adds that Muse Ember, Muse Spark and Muse Wonder are now available on Google Cloud and Oracle Cloud.Source: research.meta.ai · checked 2026-11-17
  • Meta calls Muse Ember its "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. Meta’s developer page calls it its smallest multimodal model, built for high-volume agentic workflow execution.Source: research.meta.ai · checked 2026-11-17
  • Meta says Muse Ember is the first model trained using Muse Wonder Full, and describes the lineup as Muse Wonder for the hardest problems, Muse Spark for coding and agentic workflows, and Muse Ember for repeatable coding and agentic tasks at scale. It suggests pairing Muse Wonder (plans) with Muse Ember (handles the volume).Source: research.meta.ai · checked 2026-11-17
  • Meta lists a 1M-token context window and up to 512K output tokens. It says Muse Ember natively understands text, images, audio and video, and that it supports the Responses API. Meta describes it as a reasoning model; its benchmark figures are all at max reasoning effort.Source: research.meta.ai · checked 2026-11-17
  • Meta’s post lists $0.10 per million input tokens, $0.004 per million cached input tokens and $0.50 per million output tokens for prompts up to 200K tokens; for longer prompts $0.30 input, $0.012 cached input and $1.50 output per million tokens.Source: research.meta.ai · checked 2026-11-17
  • Meta’s own figures, not tested by us (Artificial Analysis Intelligence Index v5 at max reasoning, as shown in Meta’s post): Muse Ember 36, Claude Haiku 5.5 36, GPT-6 Luna 28, Muse Spark 1.3 37; Meta puts the cost per Artificial Analysis task at about $0.14 for Muse Ember against $2.39 for Muse Spark 1.3. Meta’s own chart also shows Muse Ember behind Claude Haiku 5.5 on FrontierCode 2.0-AA (27.2 against 34.9).Source: research.meta.ai · checked 2026-11-17
  • Neither Meta’s post nor its developer page says whether the weights are released, what licence applies, or how many parameters the model has; both describe access only through the Meta Model API, Muse Code and cloud partners. We found no licence page for Muse Ember.Source: developer.meta.com · checked 2026-11-17
  • Listed on OpenRouter as meta/muse-ember with a 1,048,576-token context window, text, image, video, file and audio input and text output, at $0.10 / $0.50 per million input / output tokens, cached input $0.000004 per token (checked November 17, 2026). OpenRouter marks reasoning as mandatory, with effort levels minimal to max and a default of max; its data shows two hosts, Meta and Google (Vertex).Source: openrouter.ai · checked 2026-11-17 · second-hand

Nearest available

Models you can use now instead

$0.10 / $0.50 (per million tokens)

  • Anthropic’s model page says Claude Haiku 5.5 was released on October 7, 2026, is the latest Haiku, and will not be retired sooner than October 7, 2027. Anthropic’s launch post is dated the same day.Source: platform.claude.com · checked 2026-11-17
  • In its launch post Anthropic calls Haiku 5.5 "the cheapest, fastest, and most capable small model we’ve ever released", for high-volume, cost-sensitive tasks such as summaries, compactions, database queries and classification, and as a coding subagent alongside Opus 5.5 and Sonnet 5.5. It calls Haiku 5.5 its first Haiku-class model with an adjustable effort setting.Source: anthropic.com · checked 2026-11-17
  • Anthropic’s model page says it is built for high-volume, latency-sensitive work such as classification, routing, extraction and subagent tasks, and rates its latency "Fastest" in the current lineup (Fable 5.2 and Opus 5.5 are "Moderate", Sonnet 5.5 is "Fast").Source: platform.claude.com · checked 2026-11-17

$0.10 / $0.50 (per million tokens)

  • OpenAI: "Our most efficient model for focused, high-volume tasks".Source: developers.openai.com · checked 2026-11-11
  • Context 1,050,000 tokens; maximum output 128,000 tokens; text and image input, text output; knowledge cutoff May 18, 2026.Source: developers.openai.com · checked 2026-11-11
  • Reasoning effort can be set to none, low, medium (the default), high, xhigh or max.Source: developers.openai.com · checked 2026-11-11

FAQ

Questions

What is Meta Muse Ember?

Muse Ember is Meta’s smallest and most efficient model, released on November 9, 2026. Meta says it is designed for complex reasoning execution and the work agents repeat all day: tool calls, coding subtasks, routing, extraction and long-document parsing. You can chat with it on this page on credits, without a Meta account.

How much does Meta Muse Ember cost?

Meta lists $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 200K tokens, and $0.30 / $1.50 for longer prompts; cached input is $0.004 (up to 200K). OpenRouter lists the same $0.10 / $0.50 (checked November 17, 2026). On this site a typical reply to this model costs about 1 credit (a credit is about $0.01 in the $5 pack), billed by its real token cost, at least 1 credit.

How do I use Meta Muse Ember?

Through the Meta Model API as muse-ember-1.0 or in Muse Code; Meta also says it is available on Google Cloud and Oracle Cloud. It is on OpenRouter as meta/muse-ember, and you can use it in the chat box on this page. We could not confirm from Meta’s pages whether it is available in the Muse app.

What is the Meta Muse Ember context window?

Meta lists a 1M-token context window and up to 512K output tokens. It takes text, images, audio and video as input. Prompts over 200K tokens are billed at higher prices. It is a reasoning model, so part of the output budget can go to reasoning.

Is Meta Muse Ember open source?

We could not confirm that. Meta’s post and developer page describe access through the Meta Model API, Muse Code and cloud partners, and do not mention released weights, a licence or a parameter count. Check Meta’s own pages before assuming it can be downloaded.

How is Muse Ember different from Muse Spark 1.3 and Muse Wonder?

Per Meta: Muse Wonder is for the hardest problems, Muse Spark for coding and agentic workflows, and Muse Ember for repeatable coding and agentic tasks at scale; it is the first model trained using Muse Wonder Full. Meta’s own figures put Muse Ember at $0.10 / $0.50 against $1.25 / $4.25 for Muse Spark 1.3, with Intelligence Index scores of 36 and 37. We have not tested the difference.

How does Meta Muse Ember compare with Claude Haiku 5.5?

Both list $0.10 per million input and $0.50 per million output tokens for shorter prompts. Meta’s own figures say Muse Ember matches Haiku 5.5’s best Artificial Analysis score (36) at about a third of the cost per task, but trails it on FrontierCode 2.0-AA. These are Meta’s figures, not tested by us.

Is this the official Meta website?

No. This is an unofficial page, not affiliated with or endorsed by Meta. Facts link to Meta’s own pages where possible.