Meta 发布 Muse Ember：面向智能体工作的小型推理模型，支持 1M 上下文
Today we're introducing Muse Ember, our smallest and most efficient model, built for complex reasoni...
Meta 发了个小模型 Muse Ember，专干智能体的活：串工具调用、跑测试、修代码，还有 1M token 上下文和音视频理解。跟 OpenAI、Anthropic 的智能体模型正面竞争，做工作流自动化的开发者可以看看。
Meta 宣布推出 Muse Ember，定位为其体量最小、能效最高的模型，主打复杂推理执行与日常智能体任务。模型可以跨应用承接工作流、按顺序调用工具、运行测试并对代码做针对性修复。上下文窗口达到 1M token，并原生支持文本、图像、音频和视频四种模态的输入。
在能力设计上，Muse Ember 的重点是把推理用在执行环节，而不是单纯的对话或问答：智能体可以在多个应用之间推进任务，串联工具调用，跑测试后再修改代码，形成一套可操作的工作循环。1M token 的上下文让它能处理大型文档集合，多模态输入则覆盖了混合格式的日常工作场景。
Meta 将其定位为小型高效模型，直接对标智能体执行层的低成本需求，与 OpenAI、Anthropic 在智能体工作流方向的产品形成竞争。对开发者而言，小参数量加上长上下文和多模态，意味着部署成本和响应速度可能更有优势。
Today we're introducing Muse Ember, our smallest and most efficient model, built for complex reasoni...
Today we're introducing Muse Ember, our smallest and most efficient model, built for complex reasoning execution and everyday agent work. Muse Ember can carry workflows across apps, call tools in sequence, run tests and make targeted code fixes. With a 1M-token context window and native understanding of text, images, audio and video, it can also work across large document sets and varied inputs.📊 20 ⚡
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison and AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Meta, Muse Ember, 多模态, 智能体.
Source attribution: curated from Meta AI (@AIatMeta). Cross-source corroboration: 11 independent sources have covered this development (including kimmonismus, techcrunch, Artificial Analysis), confirming industry-wide signal strength.
