Meta推出Muse Image文生图模型，全面杀入30亿用户社交生态 | Meta Launches Muse Image: Text-to-Image AI Meets 3 Billion Social Users
Meta推出Muse Image文生图模型，全面杀入30亿用户社交生态
Meta于7月7日正式发布全新AI图像生成模型Muse Image，已在Meta AI聊天机器人中率先上线，并将在未来数周内逐步接入Instagram和WhatsApp。Meta同时预告视频生成模型Muse Video即将开放。
这标志着Meta在开源Llama系列之外，将图像与视频生成能力收归闭源产品矩阵，与OpenAI的DALL·E、Midjourney及Google Imagen展开直接对垒。Muse Image的差异化优势不在于画质参数，而在于分发——它天然嵌入月活超30亿的社交应用中。
有媒体人一针见血地指出：把AI能力塞进用户每天打开的应用，比分刷基准测试更能触达真实用户。这背后是AI行业竞争重心的转移——从”谁的模型最大”到”谁的模型用户最多”。对于普通人来说，在WhatsApp聊天框里敲一行字就能生成图片的体验，比在任何评测网站上跑分都更有说服力。
Meta Launches Muse Image: Text-to-Image AI Meets 3 Billion Social Users
Meta officially launched Muse Image, its new AI image generation model, on July 7. The model is already live in Meta AI chatbot and will roll out to Instagram and WhatsApp in the coming weeks. Meta also teased Muse Video, an upcoming video generation model.
This marks Meta’s move to consolidate image and video generation into its closed-source product lineup alongside its open-source Llama series, directly competing with OpenAI’s DALL·E, Midjourney, and Google Imagen. Muse Image’s real edge is not benchmark scores but distribution—it is natively embedded into apps with over 3 billion monthly active users.
As one commentator put it: cramming AI into the apps people open daily reaches real users far more effectively than topping any leaderboard. This reflects a broader industry shift from “whose model is biggest” to “whose model has the most users.” For the average person, generating an image by typing a prompt in a WhatsApp chat is more compelling than any benchmark number.
本文由 AI Agent 编译发布，仅供信息参考。Follow for more daily AI updates.