Meta is making a bold move in AI with two new generative models named after fruits. According to The Wall Street Journal, Meta is developing Mango, focused on image and video generation, and Avocado, a large language model aimed at enhancing text and code capabilities. Having fallen behind competitors like OpenAI, Anthropic, and Google, Meta’s existing assistant, Meta AI, remains embedded in its platforms but lacks broad traction. With Mango and Avocado, Meta aims to build foundational AI technologies rather than just functional tools. Both models are expected to debut in the first half of 2026, developed under Meta’s new superintelligence division led by Scale AI co-founder Alexandr Wang. This development is critical for marketers to watch. **What’s happening at Meta’s AI lab** Meta Superintelligence Labs (MSL), restructured with new leadership including Alexandr Wang and Chief Product Officer Chris Cox, is driving these initiatives. They aim to create “world models” that can generate content, reason, plan, and act without exhaustive training on all scenarios. Avocado is designed for coding and logical reasoning, while Mango targets high-quality image and video generation, supporting Meta’s focus on multimodal AI—which underpins everything from content creation to AR and VR experiences. However, challenges remain; despite new talent inflows, some employees have left recently, including Chief AI Scientist Yann LeCun, who is starting his own startup. **Why this matters for AI marketing** So far, Meta’s AI features have been subtle, embedded in platforms like Facebook, Instagram, and WhatsApp as search or content suggestions. Mango and Avocado represent a step change—foundation models that could fuel a new generation of creative and coding tools directly integrated into Meta’s ecosystem. This could enable auto-generated social videos, branded visuals, or campaign copy native to Meta’s apps. For marketers, this implies: - Native AI creation tools integrated into Meta’s ad and content platforms, reducing reliance on third-party AI creative tools. - New interfaces such as conversational ad creation, smart campaign planning, and automatically generated assets within Meta Business Suite or Creator Studio. More importantly, Avocado’s advanced reasoning capabilities could transform marketer interactions with AI from simply using dashboards to collaborating with AI systems. **Strategic takeaways for marketers** To prepare for Mango and Avocado’s arrival in 2026, marketers should consider: 1.

**Plan for platform-native generative AI:** Early mastery of prompt engineering and template design will help brands scale creative output as AI generation becomes embedded in platforms like Instagram and WhatsApp. 2. **Focus on content authenticity and trust:** As AI-generated content rises, brands must maintain clear disclosure practices and consistent voice to preserve consumer trust. 3. **Explore multimodal ad experiences:** Combining text, image, and video powered by Mango and Avocado could reshape dynamic ad formats. Marketers should familiarize themselves with prompt engineering and dynamic creative optimization. 4. **Anticipate disruption to AI tool ecosystems:** Successful foundation models from Meta may reduce demand for standalone generative AI platforms, requiring companies to reassess their creative technology stacks. Mango and Avocado mark a significant strategic push by Meta to regain leadership in generative AI. Success depends not only on technology but on effective execution, integration, and building trust with users. For marketers, now is the moment to audit AI workflows, rethink content pipelines, and monitor Meta’s platform strategies. Although Mango and Avocado launch in 2026, preparing early can provide a crucial competitive edge.