Home / Technology / Meta's Superintelligence Lab Fuels AI Race
Meta's Superintelligence Lab Fuels AI Race
19 Dec, 2025
Summary
- Meta is developing 'Mango' image/video and 'Avocado' text AI models.
- New AI models from Meta are slated for release in the first half of 2026.
- Meta aims to improve coding and world understanding with its new AI.
Meta is intensifying its artificial intelligence development through its Superintelligence Lab, led by Scale AI co-founder Alexandr Wang. The lab is currently focused on creating two major models: 'Mango,' an advanced image and video AI, and 'Avocado,' a new text-based AI. These cutting-edge models are scheduled for release in the first half of 2026, as revealed in a recent internal Q&A.
The company's objective with these new AI tools is twofold: to significantly improve coding proficiency and to develop 'world models' capable of understanding visual information, reasoning, planning, and acting autonomously. This initiative aims to bridge the gap Meta faces compared to competitors such as OpenAI, Anthropic, and Google in the rapidly evolving AI race.
Despite recent organizational shifts and departures within its AI division, Meta is pinning high hopes on these upcoming projects. The success of 'Mango' and 'Avocado' is crucial, as they represent the first major outputs from the Superintelligence Lab and could define Meta's future standing in the AI arena. The company's strategy involves integrating its AI assistant across its vast social networks to boost user numbers.
