Meta Platforms has reportedly delayed the launch of its next flagship artificial-intelligence model, code-named “Avocado,” after internal testing revealed performance issues compared with competing AI systems.

The model was originally expected to launch earlier in 2026 but will now likely be released later in the year as engineers work to improve its capabilities. Reports indicate the system has not yet matched the performance of leading AI models developed by companies such as Google and OpenAI. 

The delay highlights the intense global competition in artificial intelligence development, where technology companies are investing billions of dollars to build more powerful AI models and infrastructure.

Despite the setback, Meta continues to expand its AI ambitions. Chief executive Mark Zuckerberg has pledged major investments in computing infrastructure and research as the company seeks to compete with industry leaders in generative AI.

Analysts say the delay reflects the complexity of developing large-scale AI models as companies attempt to push the boundaries of machine learning and computing power.

Sources: Reuters, Investors Business Daily