Meta is developing a new image and video model for a 2026 release, report says

TechCrunch
Meta is developing new AI models, including an image/video model named "Mango" and a text model "Avocado," for a 2026 release.

Summary

Meta is actively developing new artificial intelligence models through its superintelligence lab led by Alexandr Wang. The company is working on an image and video model codenamed "Mango" and a text-based model called "Avocado," with plans to release both in the first half of 2026, according to a report citing an internal Q&A. Wang stated the text model will focus on improved coding capabilities, while Meta is also exploring world models that can understand visual data, reason, plan, and act without exhaustive training. This development comes as Meta faces pressure in the AI race against rivals like OpenAI and Google, following internal restructurings and the departure of some researchers, including chief AI scientist Yann Lecun. Since Meta currently lacks a standout AI product, the initial releases from Meta SuperIntelligence Labs (MSL) are considered highly significant for the company's standing.

(Source:TechCrunch)