Meta Delays Avocado AI Model Launch Due to Poor Performance

Meta has delayed the launch of its Avocado AI model due to underperformance compared to competitors like Google's Gemini 3.0. According to insiders, Avocado outperformed Meta's previous AI model but struggled to compete with Gemini 3.0. The delay puts Meta's $135 billion AI bet under scrutiny.

META has quietly pushed back the launch of its next-generation AI model, Avocado, from its initially planned debut this month to at least May, according to a report published on ZeroHedge News . The decision comes after internal tests showed that Avocado underperformed on key benchmarks for reasoning, coding, and writing in comparison to its rivals. According to some of the sources, Avocado trailed behind Google's Gemini 3.0, though it performed better than Meta's own prior efforts.

While the news was reported by multiple outlets, including The New York Times and Barron's , the primary details of the report are attributed to ZeroHedge News . The delay raises concerns about Meta's ability to compete at the frontier of AI development against OpenAI, Google, and Anthropic.

The setback puts Meta's massive AI investment — with capital spending plans of $115 billion to $135 billion for 2026 — under increased scrutiny. In a surprising development, the company's leadership reportedly discussed temporarily licensing Google's Gemini technology to power certain Meta products while Avocado is brought up to competitive standards. For META investors, the delay raises questions about whether Meta can close the gap with frontier AI leaders or whether its open-source AI strategy needs a fundamental rethink.