Meta is accelerating its efforts in artificial intelligence, aiming to launch advanced AI models in the first half of 2026. The company’s superintelligence lab, now led by Scale AI co-founder Alexandr Wang, is focusing on both image and video as well as text-based AI systems.
According to sources, Meta is developing an image and video model called “Mango” and a new text-based model internally named “Avocado.” The roadmap was revealed during an internal Q&A with Wang and Meta’s Chief Product Officer, Chris Cox.
Wang stated that Meta aims to enhance the text-based model’s coding abilities while exploring “world models” capable of understanding visual information, reasoning, planning, and acting without needing exhaustive training. This approach reflects Meta’s ambition to compete with AI leaders like OpenAI, Google, and Anthropic.
Meta’s AI division has undergone significant restructuring this year. Leadership changes and aggressive hiring from rival firms have been central to the strategy. However, retention has been challenging, with several new hires already departing. The situation intensified when Meta’s chief AI scientist, Yann LeCun, announced in November that he would leave to start his own venture.
Despite high usage of Meta’s AI assistants integrated into Facebook, Instagram, and WhatsApp, the company has not yet released a standalone AI product that matches the success of competitors. Analysts see the upcoming Mango and Avocado models as crucial to Meta’s reputation in AI innovation.
Meta is betting that these new tools will demonstrate its capability to lead the next generation of AI technologies. The company aims to prove that it can create AI systems that are not only intelligent but also versatile, addressing a wide range of tasks efficiently.
With the 2026 launch window approaching, all eyes are on Meta to see if it can deliver impactful AI products and establish a stronger foothold in a highly competitive market.
In other news read more about: Meta to Invest $65 Billion in AI Expansion in 2025
