July 13, 2026, 5:38 am | Read time: 4 minutes
Meta is expanding its AI offerings by integrating a new image generator, Muse Image, into its applications. The technology is designed not only to create new images but also to alter existing photos.
New Image Model from In-House AI Division
Behind Muse Image is a new model from the “Meta Superintelligence Labs.” According to the news agency Reuters, this is an AI division of the company established at the end of June 2025. According to Meta, Muse Image is the first image model from this area.
The system is intended to both generate new images and edit existing ones. Meta aims to integrate creative functions more deeply into its own platforms. Images can be created and reused directly in chats, stories, or feeds without needing additional programs.
Image Creation via Natural Language
Operation is conducted through Meta AI. Users can describe via the Meta AI app or the platform meta.ai what image they want to create or what change should be made to an existing photo.
Muse Image is also expected to handle more complex inputs. According to the company, the model can prepare multiple steps, use additional web information as context, and combine various image references. For example, a selfie and a vacation photo can be merged into a personalized postcard.
Other use cases mentioned by Meta include pets styled as paintings or enhancing a product photo based on an existing image.
Changes Directly in Conversation with AI
Not only creation but also editing should be possible directly within an AI chat. Users can mark, circle, or annotate image areas with short notes.
Meta AI is then supposed to specifically alter only the selected part. The rest of the image remains unchanged. Further adjustments can be made directly in the same conversation. For instance, styles can be changed, individual elements added, or backgrounds swapped.
According to Meta, this makes editing more like a conversation with AI than working in traditional image editing software.
Presets to Simplify Use
Additionally, Meta provides so-called presets. These are pre-made templates that can be used in just a few steps. The possibilities mentioned include refreshing old photos, trying out different hairstyles, or converting portraits into clay animation or 16-bit style.
Furthermore, Muse Image is expected to reliably display text in AI-generated images. Users could create invitations, posters, or infographics without needing to correct text afterward.
Instagram Warns Against AI Posts
Pixel Smartphones Get Theme Packs and More New Features
Close Integration with Instagram and WhatsApp
Muse Image will be closely linked with existing Meta services. On Instagram, the company already uses the model for more than 30 new AI effects in stories.
Also of interest: Meta must open WhatsApp for third-party AI
WhatsApp is also set to receive the technology. There, image generation will be available directly within conversations with Meta AI. Additionally, public Instagram profiles can be incorporated into image ideas in the Meta AI app via @-mentions. According to Meta, this allows for the creation of shared themes, invitations, or trend-related graphics.
At the same time, the company points out that users should be able to decide whether their content can be marked for such AI creations.
Availability Not Fully Known Yet
Meta has not yet disclosed when Muse Image will be available worldwide. The company only mentions selected countries where the launch will initially take place. Which regions are included has not been specified.
Later, the technology is also planned to be used on Facebook, in Messenger, and in other areas of Instagram and WhatsApp. Additionally, Meta plans to offer Muse Image to advertisers through Advantage+ Creative.