NEW YORK, April 23, 2026, 9:48 AM EDT
- Alphabet Class A was trading close to $340 early Thursday in New York, leaving the tech giant’s market cap holding above $4.1 trillion as Google leveraged Cloud Next to expand its enterprise-AI ambitions.
- Alphabet’s move comes just days ahead of its April 29 report, as the market tries to decide if cloud and AI sales will be enough to support that massive $175 billion to $185 billion capital budget slated for 2026.
- Google is rolling out AI agents, upgraded security tools, and a pair of new custom TPU chips, all part of a push to expand its cloud share against Amazon and Microsoft. Still, it’s keeping Nvidia GPU capacity in the mix.
Shares of Alphabet’s Google edged close to $340 on Thursday, following the debut of fresh enterprise AI tools and custom chips during its Cloud Next event. Investors kept eyes on Google Cloud in the leadup to earnings next week. The Class A stock changed hands at $339.89 during the morning, putting Alphabet’s valuation right around $4.1 trillion.
The calendar’s crucial here. Alphabet is set to deliver its first-quarter numbers on April 29, and investors are pressing for evidence that all the AI outlays are yielding more than just ballooning data-center expenses. Back in February, execs flagged that 2026 capex could hit $175 billion to $185 billion, and this week, Pichai noted a bit over half of this year’s machine-learning compute budget is earmarked for the cloud segment.
At its Las Vegas gathering, Google rebranded much of its enterprise AI portfolio, putting it all under the Gemini Enterprise banner. The company expanded its old Vertex AI suite, turning it into a bigger platform for developing, deploying, and managing AI agents—these are the kinds of software that handle multi-step tasks with only minimal human involvement. New security and governance tools are coming, too, targeting businesses focused on compliance and oversight.
Thomas Kurian, CEO of Google Cloud, didn’t mince words: “The experimental phase is behind us, and now the real challenge begins.” Speaking to Reuters, he noted that demand has moved away from traditional machine-learning projects and described a sudden “explosion” in custom agents. Reuters
Google rolled out hardware alongside the software push, introducing two new eighth-generation custom AI chips. The TPU 8t is built for training large models, while the TPU 8i handles inference—when the model generates a response. Customers will be able to run these chips together with Nvidia GPU instances, according to the company.
So Google is juggling multiple battles. According to Reuters, its share of the cloud market hit 14% by the end of 2025—still behind Amazon and Microsoft. Kurian, for his part, insisted Google brought “capability in the platform that nobody else offers.” Reuters
Pichai cited new usage figures: Google’s models are now handling over 16 billion tokens per minute for customers, up from 10 billion last quarter. (Tokens are the bits of text and data that AI parses.) Paid monthly active users for Gemini Enterprise climbed 40% from the prior quarter in the first quarter.
He noted that AI now produces 75% of all new code at Google, with engineers giving the green light—a jump from 50% last fall. Internally, Google points to this figure as evidence that the tools function at scale, beyond just demo environments.
Still, the cost issue lingers. Back in February, Bernstein’s Mark Shmulik flagged a warning: megacap tech could be pouring “north of a trillion dollars” into AI by 2026. For that bet to make sense, he said, the AI product market would have to multiply in size—fast. Reuters
Alphabet’s main pitch at the moment? Rapid cloud growth, a bigger AI push for business, and an earnings release right around the corner. What investors actually need to see is tougher: real revenue, margins, and proof that the conference talk turns into sales.
