Meta launches Muse Ember, its lowest-cost reasoning AI model yet
Meta has released Muse Ember, a lightweight reasoning model priced at $0.10 per million input tokens, taking direct aim at rival offerings from Anthropic and OpenAI.
Meta has introduced Muse Ember, which the company describes as its smallest, lightest, and most cost-efficient AI model yet. Built to handle complex reasoning and agent workloads, the model is available through the Meta Model API at $0.10 per million input tokens and $0.50 per million output tokens. That makes its input pricing approximately 12 times cheaper than Muse Spark 1.3 and roughly 40 times cheaper than Muse Wonder.[1][2][6][7][8][9][10]
Trained using Muse Wonder Full, the multimodal model supports text, images, audio, and video. Muse Ember is designed to execute repetitive agent tasks such as document and image processing, coding, and computer use. According to benchmark figures, Muse Ember scored 36 on the Artificial Analysis Index, matching Anthropic's Claude Haiku 5.5 and outperforming OpenAI's GPT-6 Luna, which scored 28. It also logged scores of 37.8 on AutomationBench-AA, 61.6 on Agents' Last Exam, and 46 on the Coding Agent Index.[3][4][5][7][9][11]
The release comes as Meta targets businesses looking to scale artificial intelligence deployments. According to The Information, Meta is currently building an enterprise sales team to market its models to corporate clients, collaborating with consulting firms including Accenture and Deloitte.[5][6]
Key facts
- Meta launched Muse Ember, its smallest and most cost-efficient reasoning model to date.
- Muse Ember is priced on Meta's Model API at $0.10 per million input tokens and $0.50 per million output tokens.
- Muse Ember's input token price is roughly 12 times cheaper than Muse Spark 1.3 ($1.25) and 40 times cheaper than Muse Wonder ($4.00).
- The model tied Claude Haiku 5.5 with a score of 36 on the Artificial Analysis Index, beating GPT-6 Luna's score of 28.
- Muse Ember supports text, images, audio, and video, and was trained using Muse Wonder Full.
- Meta is reportedly establishing an enterprise sales team to pitch its AI models to businesses, partnering with consulting firms such as Accenture and Deloitte.
Sources · 11 sources
- CO
Cointelegraph@CointelegraphPost on X ·
⚡️ NEW: Meta has launched Muse Ember, its cheapest AI model yet, priced at $0.10 per million input tokens. https://t.co/lDkcZc09UC
Open source - CB
Crypto BriefingArticle ·
Meta launches Muse Ember, its smallest reasoning model, at $0.10 per million input tokens Meta's affordable AI model could democratize access to reasoning capabilities, but data-sharing terms may deter privacy-conscious developers. The post Meta launches Muse Ember, its smallest reasoning model, at $0.10 per million input tokens appeared first on Crypto Briefing .
Open source - TN
The New Stack@thenewstackPost on X ·
Meta’s Muse Ember is a low-cost multimodal AI model for coding and computer use, with benchmark scores close to Claude Haiku 5.5 and Meta’s Muse Spark 1.3. https://t.co/BwFQLIkmGq
Open source - SI
SiliconANGLE@SiliconANGLEPost on X ·
Meta introduces Muse Ember to power cost-efficient AI agents https://t.co/Jpw2Iy8CB9
Open source - AI
Analytics Insight@analyticsinmePost on X ·
𝗠𝗲𝘁𝗮 𝗟𝗮𝘂𝗻𝗰𝗵𝗲𝘀 𝗠𝘂𝘀𝗲 𝗘𝗺𝗯𝗲𝗿, 𝗟𝗶𝗴𝗵𝘁𝘄𝗲𝗶𝗴𝗵𝘁 𝗔𝗜 𝗠𝗼𝗱𝗲𝗹 𝗳𝗼𝗿 𝗔𝗴𝗲𝗻𝘁 𝗘𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻 Meta is making its AI agents lighter, faster, and cheaper! Meet Muse Ember, the company's new lightweight model designed to handle repetitive tasks without the hefty costs of larger AI systems. From processing documents and images to managing thousands of operations, this model targets businesses ready to scale AI. Can cheaper AI execution change the game? #MetaAI #MuseEmber #AIAgents #ArtificialIntelligence #TechNews #AnalyticsInsight #AnalyticsInsightMagazine Read More 👇 https://t.co/0sqHYrQf3M
Open source - WS
Wall St Engine@wallstenginePost on X ·
$META JUST LAUNCHED MUSE EMBER, ITS SMALLEST AI MODEL Meta says Muse Ember is its smallest and most efficient model, built for complex reasoning. It's available today on Meta's Model API for $0.10 per million input tokens and $0.50 per million output tokens. Tokens are the small chunks of text an AI model reads and writes. That's about 12x cheaper on input than Muse Spark 1.3, which lists at $1.25, and 40x cheaper than Muse Wonder, Meta's priciest model at $4. It comes as Meta builds a sales team to sell its AI to businesses, working with consulting firms like Accenture and Deloitte, according to The Information.
Open source - TE
TechmemeArticle ·
Meta releases Muse Ember, its smallest and most cost-efficient AI model yet, saying it delivers more intelligence per dollar than Haiku 5.5 and GPT-6 Luna (Russell Brandom/TechCrunch) Russell Brandom / TechCrunch : Meta releases Muse Ember, its smallest and most cost-efficient AI model yet, saying it delivers more intelligence per dollar than Haiku 5.5 and GPT-6 Luna — Meta Superintelligence Lab released a new AI model on Monday called Muse Ember, the lab's lightest and most cost-efficient AI model yet.
Open source - SB
Shay Boloor@StockSavvyShayPost on X ·
$META just launched Muse Ember which is its smallest AI model priced at $0.10 per million input tokens and $0.50 per million output tokens making input 40x cheaper than Muse Wonder. And that’s what makes the AI price war so interesting because as intelligence gets cheaper then number of things worth automating grows and every one of those new workloads still needs compute.
Open source - DE
Decrypt@DecryptMediaPost on X ·
Meta Launches Muse Ember, Matching Claude Haiku 5.5 at a Third of the Cost. Meta's new efficient AI model scored 36 on the Artificial Analysis Intelligence Index, level with Claude Haiku 5.5, and costs $0.10 per million input tokens and $0.50 per million output tokens via the Meta Model API.
Open source - MT
MTS@MTSlivePost on X ·
SITUATION DETECTED: Meta has released Muse Ember, their smallest and most efficient model. https://t.co/LEQX5Tn7T2
Open source - CH
Chubby♨️@kimmonismusPost on X ·
First release of the day: Meta’s new Muse Ember beats Claude Haiku 5.5 and GPT-6 Luna on key agent and coding benchmarks. It matches Haiku's overall intelligence score at roughly one-third the cost per task. All three start at $0.10/$0.50 per million input/output tokens. Ember is Meta’s first model trained using Muse Wonder Full and supports text, images, audio and video. Great launch and release! Meta is super back!
Open source