Full Breakdown
Meta Delays Launch of AI Model Avocado Amid Performance Concerns
3/15/2026, 12:21:38 AM
Internal Testing Reveals Shortcomings
Meta Platforms, Inc. has postponed the release of its new artificial intelligence model, code-named "Avocado," to at least May 2026. Originally slated for a March debut, the delay follows internal testing that indicated Avocado's performance in reasoning, coding, and writing fell short compared to leading models from Google, OpenAI, and Anthropic. Specifically, Avocado outperformed Meta's previous model, Llama 4, and Google's older Gemini 2.5, but could not compete with Google’s latest Gemini 3.0, which was released in November 2025.
Heavy Investments and Strategic Shifts
Meta has committed substantial resources to its AI initiatives, projecting capital expenditures between $115 billion and $135 billion for 2026, nearly double the previous year's spending. This investment includes a $14.3 billion stake in Scale AI, which led to the appointment of Alexandr Wang as Chief AI Officer. Under Wang's leadership, Meta's AI division, TBD Lab, has faced internal challenges, including turnover among researchers and disagreements over the direction of AI development.
The company is reportedly considering temporarily licensing Google’s Gemini technology to support its AI products while Avocado is being refined. This potential move marks a significant shift in Meta's strategy, which has traditionally emphasized open-source models. The discussions around licensing Gemini signal a pivot towards proprietary systems, reflecting a response to competitive pressures in the AI landscape.
Official Statements and Responses
A Meta spokesperson stated, “As we've said publicly, our next model will be good, but more importantly, show the rapid trajectory we're on, and then we'll steadily push the frontier over the course of the year as we continue to release new models.” This sentiment echoes CEO Mark Zuckerberg's earlier assertions about the ambitious goals for Meta's AI capabilities.
Criticism and Opposition
Critics have pointed out that despite heavy investments, Meta's AI efforts have not yielded the expected results. The company has faced scrutiny over its ability to deliver competitive AI models, especially after the disappointing performance of Llama 4 and the ongoing delays with Avocado. Additionally, the consideration of licensing technology from a competitor raises questions about Meta's internal capabilities and strategic direction.
Conflicting Reports and Gaps
While Meta's internal tests indicate that Avocado is underperforming, there are no definitive reports on when the model will achieve competitive parity with its rivals. The discussions about licensing Gemini also remain unresolved, leaving uncertainty about Meta's immediate AI strategy.
What's Next for Meta's AI Ambitions
Looking ahead, Meta is reportedly developing additional models, including one codenamed "Watermelon," which is intended to succeed Avocado. However, the timeline for these developments is now less certain, given the current setbacks. As Meta continues to navigate its AI ambitions, the company faces the dual challenge of delivering competitive products while managing the financial implications of its extensive investments in AI infrastructure.
