On July 7, Meta announced the launch of Muse Image, a new artificial intelligence image generation tool, developed and launched by its dedicated AI department Meta Superintelligence Labs. The new feature, codenamed "Mango" internally at the company, is now available to users for free through the Meta AI app and is also integrated into products such as Instagram Stories and WhatsApp.
From the perspective of functional positioning, Muse Image is similar to most AI image generators currently on the market. Users can generate a large number of exaggerated, humorous or cartoon-style images through text prompts for daily social sharing and creative expression. In response to the lack of inspiration for some users in writing prompt words, Meta has added presets to Muse - pre-made image prompts to "stimulate creativity" and help users get started quickly.
Meta demonstrated a variety of application scenarios in the accompanying demo video, including quickly generating customized advertising creatives through Muse and virtually trying out interior decoration solutions. In the video, a user uses Muse to simulate the effect of placing a second-hand sofa in his or her garage. Such functions will be deeply integrated with Facebook Marketplace, a second-hand trading platform owned by Meta, to improve the display efficiency of second-hand furniture and household items.
In addition to generating images from scratch, Muse Image also supports image editing functions based on prompt words, which facilitates users to re-create content within the Meta ecosystem and distribute it to different platforms. The official introduction mentioned that users can ask the system to generate a composite photo of themselves standing in front of a certain historical landmark, or to "cleanly" erase passers-by who intruded into the shot with one click, or to generate a functional QR code with customized prompts that can be actually used.
Launching alongside Muse is a new set of AI effects based on the model, focused on Instagram Story scenes. These special effects are mainly customizable filters, which support stylization and visual enhancement of existing photos, further enriching creators' expression methods in short content.
In terms of the charging model, Meta said that Muse Image will remain free for the "daily creation" stage, but when usage exceeds a certain threshold, users will need to continue using the service through a subscription plan. The company also revealed that Muse Video for video content has entered the development stage and is generally regarded as the next AI video generation product to be launched; TechCrunch has initiated further inquiries to Meta for follow-up details.
In the past year, Meta has intensively released a number of AI-related applications and services, including Creator, an AI assistant launched on the Facebook platform, and Pocket, an application that can be used for "vibe-style" code creation to help users create video games. Although the outside world has been questioning Meta's overall AI strategy as "ambiguous", the company still plans to continue to invest huge amounts of money to build AI infrastructure this year to support the expansion and operation of these new services.