Rohan Paul· @rohanpaul_ai · X·· 2026-07-12AI 评分44
AI 导读
Meta 的 Muse Spark 1.1 的多模态推理能力。它从手机获取视频，推理提取照片和产品详情，然后通过计算机使用操作浏览器，填写 Marketplace 表单、上传图片并发布商品列表。
正文 · 原文
Meta's Muse Spark 1.1 's multimodal reasoning ability.
Here it took videos from mobile, reaons to extract photos & product details then operates the browser with computer use actions to fill Marketplace forms, upload images and publish listings.
来源：Rohan Paul · x.com