Meta has officially unveiled Muse Image, its first dedicated AI image generation model developed by its Superintelligence Labs division. Announced on Tuesday, the model will immediately power image creation tools across the Meta AI app, Instagram, and WhatsApp, with expansions to Facebook and Messenger scheduled for the near future. The launch marks a strategic shift for the company, as the new Muse family of models begins to replace the previously established Llama lineup under the direction of division head Alexandr Wang. Unlike conventional generative tools, Muse Image operates with agentic capabilities by integrating with the Muse Spark large language model. This architecture enables the system to analyze user prompts, conduct web searches, and formulate a generation plan before rendering final outputs. The update also introduces a highly social feature allowing users to tag Instagram accounts directly within their prompts. When a username is included, the model accesses public imagery to incorporate recognized likenesses into generated compositions. Meta emphasizes that users retain granular control over how their content is utilized for AI purposes, with privacy and consent settings available across the platform. Beyond social tagging, the model offers a suite of creative utilities. Users can transform existing images through suggested prompts, generate layout templates for invitations and postcards, and redesign interior spaces using reference photos sourced from Facebook Marketplace or external websites. Direct canvas editing is also supported, enabling precise adjustments through hand-drawn overlays that can be immediately shared to feeds, stories, or messaging threads. The initial deployment focuses on enhancing Instagram Stories short-form video experience. Meta will roll out thirty new AI-powered effects to Instagram Stories in the United States first, with geographic and functional expansions planned for subsequent phases. Looking ahead, the company is preparing to introduce Muse Video, a forthcoming model that Wang previewed as a direct competitor in the generative video space. Early testing indicates strong performance in prompt adherence, visual fidelity, and temporal consistency across generated clips. The integration of Muse Image across Meta core applications represents a significant consolidation of its generative AI infrastructure. By unifying image and video capabilities under a single architectural family and emphasizing social interoperability, Meta is positioning its ecosystem as a primary destination for user-driven content creation. The rollout strategy, which prioritizes phased geographic expansion and privacy-conscious social features, aligns with the company broader objective of embedding agentic AI directly into everyday digital interactions.