Meta’s Superintelligence Lab launched a pair of recent state-of-the-art fashions on Monday, Muse Surprise and Muse Surprise Full, representing the corporate’s first try at a frontier mannequin.
The place Muse Spark aimed to be a light-weight mannequin that might carry out at low latency, the Surprise fashions are bigger and mentioned to be extra succesful, with increased benchmark scores and higher outcomes on complicated duties like coding.
In response to Meta’s testing (offered in the mannequin card), Surprise and Surprise Full obtain comparable benchmark scores to Fable 5.1 and GPT-6 Astra.
“Collectively, they mark the following step on our scaling ladder and a serious leap towards private superintelligence,” Meta’s weblog submit concerning the fashions reads.
That scale is mirrored within the pricing: Accessed by means of the Meta Mannequin API, Surprise will price $4 per million enter tokens and $20 per million output tokens, making it roughly equal in price to GPT-6 Sol or Opus 5.5.
Notably, Surprise Full will solely be obtainable to Muse Code subscribers and won’t be served by means of the API. Each fashions may even be used to energy the Muse agent.
In line with Meta’s curiosity in creating private assistants that may carry out duties over a protracted interval, the corporate positioned explicit emphasis on the brand new fashions’ capability to retain and synthesize data given by the consumer, a property Meta calls “energetic reminiscence.”
To realize higher recall, Meta highlighted the MemoryBench benchmark, which challenges continuous studying and reminiscence building frameworks. In response to the corporate’s testing, Muse Surprise now scores increased on MemoryBench than Fable, Opus, or Astra.
“We put an amazing quantity of effort into ensuring your mannequin can keep in mind what’s vital to you over days, weeks, months, and past,” mentioned analysis scientist Kavya Kopparapu in a video accompanying the launch.
The brand new fashions are the primary frontier mannequin launch since Scale AI founder Alexandr Wang restructured Meta’s AI efforts in August of final yr. Since then, Meta has launched its light-weight Muse Spark mannequin and Muse agent however has averted the type of state-of-the-art mannequin releases OpenAI and Anthropic are identified for.
In response to the submit, the fashions have been educated at Meta’s Prometheus information middle in central Ohio, which got here on-line earlier this yr.
Whenever you buy by means of hyperlinks in our articles, we could earn a small fee. This doesn’t have an effect on our editorial independence.

![AI electronic mail topic traces that drive 3x extra income and truly convert [+ exclusive insights]](https://blog.aimactgrow.com/wp-content/uploads/2025/11/ai-email-optimization-1-20251014-4500151-1.webp-120x86.webp)







