WhatsApp is reportedly developing a new generative AI feature that will enable users to create personalized avatars for various imaginative settings. This feature, currently in testing within the WhatsApp Beta for Android 2.24.14.7, combines user-uploaded images, text prompts, and Meta’s AI Llama model to generate custom images.
According to a screenshot shared by WABetaInfo, users can envision themselves "in any setting from the forest to outer space." The generated avatars resemble those typically created by AI generators, similar to offerings from apps like Lensa AI or Snapchat’s “Dreams” feature.
To create these personalized avatars, users will need to provide photos of themselves, which will train Meta AI to produce images in their likeness. Users can generate their avatars by typing “Imagine me” with a description in the Meta AI chat or in any WhatsApp conversation by using “@Meta AI imagine me...” This feature will be optional, requiring users to enable it manually in their WhatsApp settings. Users will also have the ability to delete reference images at any time through the Meta AI settings.
Currently, it is unclear when this new avatar feature will be available to all users, as WhatsApp is still rolling out support for its Meta AI chatbot and broader AI image generation capabilities for users in the U.S. Meta is expected to proceed cautiously with this feature, considering past issues with its generative AI tools.