Mark Zuckerberg’s Meta has reportedly delayed the release of Avocado AI, its frontier artificial intelligence model. According to a New York Times report published on March 12, 2026, the AI model won’t come before May 2026.
The report also cited three people who highlighted that they were familiar with the latest development. Though the launch of Avocado AI had been targeted for March 2026, it is now being pushed further for other priorities.
Avocado has faced multiple delays over the past few years. Meta insiders initially expected the model before the end of 2025, but it slipped to the first quarter of 2026 amid training and performance-testing challenges, as CNBC first reported in December 2025.
Bloomberg reported around the same time that Avocado was expected to debut "sometime next spring" and could be launched as a closed, proprietary model — a departure from Meta's previous open-source Llama strategy.
The delay of Avocado AI is the latest setback in Meta's efforts to catch up with other AI competitors. Mark Zuckerberg’s AI rivals Google, OpenAI, and Anthropic have long been ahead.
In early February 2026, according to an internal memo, the Avocado AI model had the potential to outperform every other AI model. This memo came from Megan Fu, product manager at Meta Superintelligence Labs. Fu described Avocado AI as “Meta's most capable pre-trained base model to date.” The AI model promises to be “competitive with leading post-trained models even before undergoing post-training refinement.” However, despite this potential, the model remains under development.
The tech giant also plans to temporarily licence Google's Gemini technology to address this setback. The aim is to upgrade and strengthen its existing AI products and to develop Avocado AI.
Also read: Mark Zuckerberg’s Meta Acquires Moltbook; Aims to Bring on AI Agents