This effort is part of Meta’s attempt to narrow the gap with major rivals such as OpenAI, Anthropic, and Google.
-
Web Desk
-
Meta is putting its full attention on building new AI systems through its superintelligence lab, which is now headed by Scale AI co-founder Alexandr Wang.
According to The Wall Street Journal, the company is developing an image and video AI model called “Mango,” along with a text-based model known internally as “Avocado.”
Meta plans to release both models in the first half of 2026. This timeline was shared during an internal question-and-answer session on Thursday by Wang and Meta’s Chief Product Officer, Chris Cox.
Wang said the company wants to strengthen the text model’s ability to write and understand code, while also working on advanced “world models.” These models are designed to process visual information and have the ability to reason, plan, and take action without being trained on every possible situation.
This effort is part of Meta’s attempt to narrow the gap with major rivals such as OpenAI, Anthropic, and Google.
Earlier this year, Meta reorganized its AI division, making leadership changes and aggressively hiring researchers from competing companies. However, keeping talent has been challenging, as several researchers who joined Meta Superintelligence Labs have already left.
Adding to the setback, Meta’s chief AI scientist, Yann LeCun, announced in November that he would leave the company to start his own venture.
Although Meta’s AI assistants are widely used due to their integration into Facebook, Instagram, and WhatsApp search features, the company has not yet introduced a highly successful standalone AI product.
As a result, expectations are high for the first projects from Meta Superintelligence Labs, as the company aims to show it can play a leading role in the next phase of AI innovation.
