Meta’s highly anticipated next-generation AI model, internally known as “Avocado,” has been delayed again, reportedly until May. This setback underscores the intense competition in the rapidly evolving AI landscape, where companies like Google, OpenAI, and Anthropic are setting a relentless pace.
Performance Shortfalls
According to sources cited by The New York Times, Meta’s model is currently underperforming in key areas such as reasoning, coding, and writing when benchmarked against its competitors. Despite substantial investment—including a $14.3 billion stake in Alexandr Wang’s AI startup—the model has yet to meet internal standards.
Massive Investment, Lagging Results
Meta has dramatically increased its AI spending, projecting a rise from $72 billion last year to between $115 billion and $135 billion this year. This financial commitment has not translated into comparable gains in model performance, with Google’s Gemini 3, OpenAI’s GPT-5, and Anthropic’s Claude Code and Cowork consistently outpacing Meta’s development.
Strategic Moves Amid Competition
While internal development lags, Meta is simultaneously making strategic acquisitions. Most notably, the company is purchasing Moltbook, a social media platform designed exclusively for AI bots. This move suggests Meta is hedging its bets by focusing on AI-driven social applications while working to improve its foundational models.
Meta’s repeated delays highlight the difficulty of competing at the cutting edge of AI. The company faces a critical juncture: either accelerate development or risk falling further behind industry leaders.
The AI arms race is not slowing down, and Meta’s position in the market will depend on its ability to close the performance gap and capitalize on new opportunities in the AI space.
