Alphabet Shares Slip as Gemini Concerns Eclipse Revenue Beat

Alphabet’s valuation rests increasingly on whether Google can convert its vast AI infrastructure, proprietary TPUs and research base into frontier models that defend Search and expand Cloud. Researchers argue that internal competition for scarce computing capacity has constrained Google DeepMind’s ability to train and iterate Gemini, particularly in coding, while OpenAI and Anthropic have advanced. That makes model performance a strategic issue: lagging at the LLM frontier could weaken enterprise adoption and force investors to trim the premium attached to Alphabet’s AI ambitions.

Alphabet reported second-quarter results on July 22, 2026, with revenue rising 24% to $119.8 billion, above the $117.1 billion FactSet consensus, while Google Cloud sales jumped 82%. Shares nevertheless fell more than 3% after the earnings call as investors weighed compute constraints, the delayed Gemini 3.5 Pro and a higher 2026 capital-spending plan of $195 billion to $205 billion. Google said Gemini 3.5 Pro remained in testing and that it would use more third-party capacity as a bridge while expanding its own infrastructure.