Meta is reportedly accelerating its artificial intelligence development, with its Superintelligence Lab working on new image, video, and text-based AI models slated for release in the first half of 2026. According to The Wall Street Journal, these ambitious projects include an image and video model internally codenamed "Mango," alongside a new text-based model known as "Avocado."
The roadmap for these advanced AI systems was reportedly unveiled during an internal Q&A session at Meta on Thursday by Alexandr Wang, co-founder of Scale AI and head of Meta's Superintelligence Lab, and chief product officer Chris Cox.
Wang emphasized that the "Avocado" text model is being developed to excel in coding tasks. Furthermore, the lab is exploring "world models" capable of understanding visual information, reasoning, planning, and acting autonomously without requiring training on every conceivable scenario.
This aggressive push comes as Meta strives to regain ground in the competitive AI landscape, where it has recently lagged behind rivals such as OpenAI, Anthropic, and Google. The company's AI division has undergone significant restructurings this year, including leadership changes and efforts to recruit top researchers. Despite these recruitment drives, several researchers who joined Meta Superintelligence Labs (MSL) have reportedly already departed the company.
Adding to these challenges, Meta's chief AI scientist, Yann LeCun, announced his departure last month to establish his own startup.
Currently, Meta has yet to launch a truly "winning" standalone AI product. Its existing Meta AI assistant's usage figures are largely bolstered by its integration into the search bars of Meta's vast social networks, which serve billions of users.
Consequently, the success of these initial projects and models emerging from Meta Superintelligence Labs will be crucial for the company's future standing in the rapidly evolving AI sector.
