Meta has outlined plans to introduce two new artificial intelligence models, codenamed Mango and Avocado, in 2026 as part of its broader effort to expand and diversify its generative AI portfolio. The announcement signals Meta’s continued investment in large-scale AI systems as it competes with other technology companies in shaping the next phase of AI development.
According to the company, Mango and Avocado are expected to build on Meta’s existing family of AI models while targeting more specialised use cases. Although detailed technical specifications have not been disclosed, the models are positioned as part of a long-term roadmap focused on improving performance, efficiency and adaptability across a range of applications.
Meta has been steadily increasing its focus on generative AI across its products and platforms. Over the past year, the company has introduced AI features across social media, messaging and advertising tools. These include AI-driven content creation, automated moderation and personalised recommendations. The planned launch of Mango and Avocado reflects Meta’s intent to deepen its foundational AI capabilities rather than relying solely on incremental updates to existing models.
The development of multiple AI models under different codenames suggests a modular approach to AI strategy. Instead of a single, general-purpose system, Meta appears to be exploring a portfolio of models optimised for different tasks or operating conditions. This approach mirrors broader industry trends where companies are developing specialised models to balance performance, cost and scalability.
Meta executives have previously stated that future AI systems must be more efficient and accessible. As generative AI adoption grows, concerns around computing costs and energy consumption have become more prominent. Newer models are expected to focus not only on intelligence but also on resource efficiency, allowing deployment across a wider range of devices and platforms.
The planned timeline of 2026 indicates that Meta is taking a measured approach to development. Rather than rushing new models to market, the company appears focused on longer-term research and testing. This strategy may help ensure stability, safety and alignment with regulatory expectations as AI governance frameworks continue to evolve globally.
Industry observers note that Meta’s AI roadmap reflects intensifying competition among major technology firms. Companies such as OpenAI, Google and others are investing heavily in next-generation models that promise improved reasoning, contextual understanding and multimodal capabilities. Meta’s planned launches suggest it aims to remain competitive by advancing its own research while leveraging its vast user ecosystem.
Mango and Avocado are expected to support Meta’s broader ambitions across social platforms, virtual and augmented reality and advertising. AI plays a central role in how Meta personalises content, manages large-scale interactions and delivers targeted advertising. More advanced models could enhance these capabilities while enabling new features and experiences.
From an advertising and marketing perspective, improved AI models may offer more sophisticated tools for campaign creation, optimisation and measurement. As brands increasingly rely on AI to manage complex digital ecosystems, Meta’s ability to offer robust and flexible AI infrastructure could influence its attractiveness as an advertising platform.
The announcement also highlights Meta’s continued emphasis on open and scalable AI development. While details remain limited, Meta has previously supported open research initiatives and collaboration within the AI community. Future models may reflect a balance between proprietary innovation and broader ecosystem engagement.
Regulatory considerations are likely to shape how Mango and Avocado are developed and deployed. Governments and regulators worldwide are paying closer attention to AI systems, particularly around transparency, data use and safety. Meta has indicated that responsible AI practices are a priority, suggesting that governance frameworks will be integrated into model development.
The naming of the models follows Meta’s practice of using internal codenames during development. While the final branding may differ, the use of distinct names signals clear differentiation within the company’s AI strategy. This differentiation could allow Meta to tailor models for specific internal teams or external partners.
As AI systems become more deeply integrated into consumer and enterprise products, expectations around reliability and trust continue to rise. Companies are under pressure to demonstrate that new models deliver tangible improvements without introducing new risks. Meta’s extended development timeline may reflect an effort to address these expectations.
The planned launch of Mango and Avocado also aligns with Meta’s broader investment in AI infrastructure. The company has committed significant resources to building data centres and computing capabilities to support large-scale model training. These investments are essential to sustaining long-term AI innovation.
While Meta has not specified which products will first incorporate the new models, observers expect applications across social platforms, messaging services and immersive technologies. AI is increasingly seen as a unifying layer across Meta’s ecosystem, supporting both engagement and monetisation.
The announcement comes at a time when public discourse around AI is shifting from novelty to impact. Stakeholders are evaluating how AI affects jobs, creativity and information integrity. Meta’s next generation models will be developed against this backdrop, making responsible deployment a key consideration.
As 2026 approaches, further details are expected to emerge around Mango and Avocado’s capabilities and intended use cases. For now, the announcement underscores Meta’s long-term commitment to advancing artificial intelligence as a core component of its business strategy.
The launch of these models will be closely watched by industry peers, advertisers and regulators alike. Their performance and adoption could influence how AI-driven platforms evolve in the coming years and how companies balance innovation with accountability.