Meta 发布小型模型 Muse Ember，专为智能体重复性工作设计
原标题：Meta launches Muse Ember, a small AI model built for the repetitive work agents actually do
AI 摘要
Meta 发布其迄今最小、最高效的 AI 模型 Muse Ember，面向 AI 智能体日常重复性任务，如调用工具、小型编码、请求路由、数据抽取和长文档处理。Meta 称其在各推理级别上每美元智能高于 Claude Haiku 5.5 和 GPT-6 Luna，Artificial Analysis 测试中每任务成本约 0.14 美元。该模型支持文本、图像、音频和视频，单次可处理最多一百万 token，现已通过 Meta Model API 和 Muse Code 提供。
正文节选
Meta launches Muse Ember, a small AI model built for the repetitive work agents actually do Meta has introduced Muse Ember, its smallest and most efficient AI model to date. It's designed for the routine tasks AI agents handle all day: calling tools, handling small coding tasks, routing requests, extracting data, and processing long documents. Meta says Muse Ember delivers more intelligence per dollar at every reasoning level than Claude Haiku 5.5 or GPT-6 Luna. It costs about $0.14 per task in