Google's parent company saw its stock drop over 4% after reports that its Gemini 3.5 Pro AI model missed its launch window, raising questions about whether Big Tech's massive AI bets are paying off.
Alphabet promised its next-generation AI model would be ready by late June. It’s mid-July, and the model still isn’t here.
A Bloomberg report on July 16 revealed that the launch of Gemini 3.5 Pro has been delayed after developers found the model’s coding performance fell short of expectations during training. The news sent Alphabet’s stock tumbling more than 4% in after-hours trading, adding fresh anxiety ahead of the company’s second-quarter earnings report scheduled for July 22.
The $190 billion question #
Here’s the thing about Alphabet’s AI ambitions: they are staggeringly expensive. The company’s capital expenditure guidance for 2026 sits at $180 to $190 billion. For context, that’s roughly double the $91.45 billion it spent in 2025.
CEO Sundar Pichai had previously flagged supply-chain issues, power constraints, and compute limitations as obstacles to scaling Alphabet’s AI infrastructure.
The July 22 earnings call will need to offer concrete timelines and monetization roadmaps. Investors watching the call should pay close attention to two things: whether Alphabet provides a revised timeline for Gemini 3.5 Pro, and whether the company adjusts its capex guidance downward or doubles down on the $180 to $190 billion figure.
The AI arms race gets messier #
OpenAI and Anthropic have both been aggressively shipping competitive models. Every week that Gemini 3.5 Pro remains unreleased is a week where enterprise customers and developers deepen their integrations with rival platforms.
The disappointing training results that triggered the delay are particularly concerning. Coding performance is one of the most commercially valuable benchmarks for large language models, driving adoption among software developers and enterprises automating technical workflows.
What this means for crypto-native AI plays #
There’s no direct crypto angle in Alphabet’s Gemini delay. No tokens were mentioned. No protocols are involved.
Pichai’s own admissions about power constraints and compute limitations are exactly the problems that decentralized GPU networks and distributed training protocols claim to address.
When Alphabet drops 4% after hours, it drags broader tech sentiment with it. And tech sentiment, for better or worse, still correlates meaningfully with risk-asset appetite across crypto markets.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our