跳到正文
原文
Cointelegraph· Cointelegraph by Felix Ng·· 6 天前精选AI 评分65

Meta 发布 Muse Wonder AI 模型,挑战 OpenAI 和 Anthropic

Meta claims new Muse Wonder AI models rival OpenAI, Anthropic

AI 导读

Meta 发布了 Muse Wonder 和 Muse Wonder Full 两种新的 AI 模型,称其为迄今为止最强大的模型,并引入了 Active Memory 和 Active Focus 两种新能力。标准 Wonder 模型每百万输入 token 收费 4 美元,而 Wonder Full 仅对订阅 Muse Code 的用户开放。

推荐理由

原文给出了能力变化和开放入口,读者可以据此判断它会怎样改变现有工作流。

原文

Meta has unveiled two new artificial intelligence models, Muse Wonder and Muse Wonder Full, which it called its most capable models to date and a step toward “personal superintelligence.”

Meta on Monday said the models introduce two capabilities Meta calls Active Memory and Active Focus. The first carries relevant information across conversations, and the second helps the agent sustain work on difficult tasks.

Independent evaluation platform Artificial Analysis scored Wonder Full at 56 on its Intelligence Index and the standard Wonder at 52, both at maximum reasoning effort.

That puts Full above Claude Fable 5.1 and GPT-6 Astra, which each scored 53, but below Claude Opus 5.5’s 58.

Bindu Reddy, CEO of Abacus.AI, which conducts its own model evaluations, challenged the performance claims, writing on X that Muse Wonder was “nowhere near Fable or Astra level” and said detailed benchmark results would follow.

Meta’s new models are its most expensive to date, with the standard Wonder model charging $4 per million input tokens, matching Claude Opus 5.5 but undercutting GPT-6 Astra and Claude Fable 5.1, which charge $10 and $50 respectively.

Wonder Full is available only to people who subscribe to Muse Code’s $50 a month “power usage” or $200 a month “maximum usage” tiers.

Related: Global watchdog flags stablecoin risks, AI threats ahead of G20

Cointelegraph is committed to independent, transparent journalism. This news article is produced in accordance with Cointelegraph’s Editorial Policy and aims to provide accurate and timely information. Readers are encouraged to verify information independently.

来源:Cointelegraph · cointelegraph.com