# Meta — Introducing Our Smallest, Most Efficient Model: Muse Ember

- Company: Meta (meta.com)
- Announced: 2026-11-09
- Category: new-model
- Coverage: 18 outlets
- Announcement: yes
- Group: models
- Source: https://research.meta.ai/blog/muse-ember
- Record: https://forck.live/items/33534-introducing-our-smallest-most-efficient-model-muse-ember
- Subject: Muse
- Models affected: Muse Ember, Muse Wonder, Muse Wonder Full, Muse Spark, Muse Spark 1.3, Claude Haiku 5.5, GPT-6 Luna
- Context window: 1M-token context window
- Pricing: $0.10 per million input tokens, $0.004 per million cached input tokens and $0.50 per million output tokens

Meta introduced Muse Ember, its smallest and most efficient model, built for the tool calls, coding subtasks, routing, extraction and long document parsing that agents repeat all day. Meta says Muse Ember delivers more intelligence per dollar than Claude Haiku 5.5 or GPT-6 Luna at any reasoning setting, matching Haiku's best score at about a third of the cost and beating every Luna setting. The model is available today on the Meta Model API and Muse Code as muse-ember-1.0, priced at $0.10 per million input tokens, $0.004 per million cached input tokens and $0.50 per million output tokens. It carries a 1M-token context window and up to 512K output tokens, and supports the Responses API. Muse Ember natively understands text, images, audio and video, so agents can read screenshots, charts and scanned documents or work from voice memos and meeting recordings without a separate model. Meta says it evaluated Muse Ember under its Meta Superintelligence Scaling Framework before release, including chemical and biological, cybersecurity and loss-of-control risks, and found it remained below the framework's high-risk thresholds. Muse Ember is the first model trained using Muse Wonder Full, and Meta positions it alongside Muse Wonder for the hardest problems and Muse Spark for coding and agentic workflows, at about $0.14 per Artificial Analysis task.

## Evidence

Verbatim from https://research.meta.ai/blog/muse-ember:

> 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. It also sits on the efficiency frontier of the full Artificial Analysis leaderboard: no model scores higher for less per task.

## Around this story

Outlets this record can name and link:

- The Verge — Meta is launching its “smallest and most efficient” AI model: Muse Ember.: https://theverge.com/tech/1029979/meta-is-launching-its-smallest-and-most-efficient-ai-model-muse-ember
- VentureBeat: https://venturebeat.com/technology/metas-proprietary-muse-ember-nearly-matches-claude-haiku-5-5-at-1-3-the-estimated-per-task-cost
- Digital Trends: https://digitaltrends.com/computing/meta-launches-muse-ember-a-cheaper-ai-model-built-for-repetitive-tasks
- Mashable — What is Muse Ember? Meet Meta’s new low-cost AI model: https://mashable.com/tech/meta-muse-ember-ai-model
- Livemint — Meta takes on OpenAI and Anthropic with Muse Ember AI model: Features, benchmarks and pricing | Mint: https://livemint.com/ai/meta-takes-on-openai-and-anthropic-with-muse-ember-ai-model-features-benchmarks-and-pricing-11794289940494.html
- Moneycontrol — Meta introduces Muse Ember AI model, claims higher intelligence at lower cost: https://moneycontrol.com/technology/meta-introduces-muse-ember-ai-model-claims-higher-intelligence-at-lower-cost-article-14076011.html
- Neowin: https://neowin.net/news/metas-new-muse-ember-takes-on-gpt-6-luna-and-claude-haiku-55-at-a-fraction-of-the-cost
- Digit — Meta introduces Muse Ember AI model, says it performs better than Claude Haiku 5.5 and GPT 6 Luna: https://digit.in/news/general/meta-introduces-muse-ember-ai-model-says-it-performs-better-than-claude-haiku-55-and-gpt-6-luna.html
- The Deep View — Meta’s Muse Ember wins on cost of agentic tasks: https://thedeepview.com/articles/meta-s-muse-ember-wins-on-cost-of-agent-tasks
- MadRobot — Meta says its tiny new Muse Ember AI does Claude Haiku’s job for a third of the cost: https://madrobot.blog/2026/11/09/meta-muse-ember-small-model-ai-agents
- Simon Willison's Weblog — Introducing Our Smallest, Most Efficient Model: Muse Ember: https://simonwillison.net/2026/Nov/10/muse-ember
- Gigazine — Metaが小型AIモデル「Muse Ember」を発表、Claude Haiku 5.5と同等の性能を約3分の1のコストで実現: https://gigazine.net/news/20261110-meta-muse-ember

## Follow-ups

- SiliconANGLE (reported, 10 Nov): Meta introduces Muse Ember to power cost-efficient AI agents — https://forck.live/items/33775-meta-introduces-muse-ember-to-power-cost-efficient-ai-agents
- TestingCatalog (reported, 9 Nov): Meta launches Muse Ember for low-cost agentic tasks — https://forck.live/items/33698-meta-launches-muse-ember-for-low-cost-agentic-tasks
- The Decoder (reported, 9 Nov): Meta launches Muse Ember, a small AI model built for the repetitive work agents actually do — https://forck.live/items/33617-meta-launches-muse-ember-a-small-ai-model-built-for-the-repetitive-work-agents
- The New Stack (reported, 9 Nov): Meta launches Muse Ember to take on Claude Haiku and GPT-6 Luna — https://forck.live/items/33549-meta-launches-muse-ember-to-take-on-claude-haiku-and-gpt-6-luna
- TechCrunch (reported, 9 Nov): Meta releases Muse Ember, a new lightweight model for agent execution — https://forck.live/items/33542-meta-releases-muse-ember-a-new-lightweight-model-for-agent-execution
- RuntimeWire (reported, 9 Nov): Meta launches Muse Ember and prices it at $0.10 per million input tokens for agent volume work — https://forck.live/items/33665-meta-launches-muse-ember-and-prices-it-at-0-10-per-million-input-tokens-for

---

Record: https://forck.live/items/33534-introducing-our-smallest-most-efficient-model-muse-ember
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md
