Alphabet (GOOGL.US) welcomes a stunning opportunity on Wall Street: The curtain will rise on Alphabet Inc. Class CI/O 2026, and the seven major battle lines will determine the valuation of "full-stack AI".

On May 20th, 1 am Beijing time, the Alphabet Inc. Class CI/O 2026 Developer Conference will officially open in California. Unlike previous years, the coordinates of this year's I/O are particularly special - 18 months ago, this search giant was still being asked by the market whether it had missed the wave of generative AI. Today, Alphabet (GOOGL.US) stock price has surged by about 140% in the past year, reaching a 52-week high of $406.29 during intraday trading on the day before the I/O opening (May 18). Behind the soaring stock price is Wall Street conducting a "full-stack AI" reevaluation of Alphabet using real gold and silver. Plexo Capital's Founding Managing Partner and early investor in Anthropic, Lo Toney, commented, "Alphabet Inc. Class C may be the most capable company of achieving large-scale commercialization of artificial intelligence because it nearly controls every layer of the entire technology stack. We have never seen a company truly achieve such complete vertical integration from top to bottom, supporting the development of artificial intelligence." All of this AI layout is backed by astronomical levels of capital investment. Alphabet has raised its full-year capital expenditure guidance range to $180 billion to $190 billion for 2026, nearly doubling from $91.45 billion in 2025, and expects further substantial growth in capital expenditures in 2027. Capital expenditures for the first quarter reached $35.7 billion, leading to a 46.63% year-over-year drop in free cash flow. Such significant spending has raised concerns in the market about the investment return period. However, considering the $462 billion backlog of orders providing abundant future revenue visibility, as well as the explosive growth of AI product revenue by nearly 800% year-on-year, most analysts remain optimistic. Bank of America Corp analyst Justin Post reaffirmed his "buy" rating and $430 target price on the eve of I/O, suggesting that potential "surprises at the AI level" may be needed to further expand valuation multiples. Berkshire Hathaway's current CEO Greg Abel increased the company's stake in Alphabet by over 200% in his first quarter after taking office, seen as a heavyweight endorsement by large institutions. However, a high stock price also means high expectations. What investors are truly concerned about is no longer "what models Alphabet Inc. Class C has released", but whether AI can become a real growth engine for search, advertising, shopping, cloud computing, and enterprise software. Around this core proposition, the following seven key areas will be the focus of this year's I/O Conference. The next step for Gemini: The ecological logic behind version numbers The biggest suspense at the I/O conference is whether Alphabet Inc. Class C will release the next-generation flagship model, Gemini 4.0. According to forward-looking reports from multiple media outlets, Gemini 4.0 aims to achieve a significant improvement in logical reasoning capabilities, benchmarking against OpenAI's GPT-5.5 in all aspects. At the same time, a native multimodal version called "Gemini Omni" has also been frequently mentioned, rumored to able to directly generate and process multidimensional information without the need for external video or audio tools. Beta users have reported that its video generation capabilities are extremely powerful, but the computational power is also incredibly high - with some users saying that generating a short video consumed 86% of the daily quota of the AI Pro plan. However, analysts from institutions such as Citigroup Inc. point out that according to Gemini's model with a release cycle of approximately 3 to 4 months (the latest was Gemini 3.1 Pro in February 2026), it is more likely to see Gemini 3.2 or 3.5 versions at I/O, with a relatively low probability of Gemini 4.0 appearing. Citigroup maintains a "buy" rating on Alphabet with a target price of $447, emphasizing that for investors, the "model number itself is not a determining factor" - what is more important is whether the Gemini ecosystem can continue to expand and enter more core services. User growth data for Gemini supports this ecosystem expansion: as of the first quarter of 2026, paid subscribers to Alphabet Inc. Class C product suites have reached 350 million; Gemini Enterprise's paid monthly active users have grown by 40% compared to the previous month; Gemini app's monthly active users in the United States increased by 127% year-on-year in April; the token volume processed per minute by API calls has exceeded 16 billion, a 60% increase from the previous quarter. AI Agents: From chat tools to "office colleagues" If there is a consistent theme at this year's I/O, it must be "Intelligent Agents." At the Google Cloud Next conference in April, Alphabet Inc. Class C officially shifted its enterprise strategic focus to "Agent-Based AI," releasing the Gemini Enterprise Agent Platform, adding "Memory Bank" and "Memory Profile" features that allow AI agents to remember past interactions, and launching "Agent Simulation" for pre-deployment testing. Alphabet Inc. Class C CEO Sundar Pichai revealed that as of April, 75% of new code at Alphabet Inc. Class C was generated by AI and reviewed and approved by engineers. Of note, insiders revealed that Alphabet Inc. Class C is developing a 24/7 "personal agent" for Gemini, aiming to position Gemini as the operational layer of Alphabet Inc. Class C products, capable of understanding context and proactively executing tasks. In the field of intelligent agent programming, Gemini is directly competing with Anthropic's Claude Code and OpenAI's Codex, a category that has become one of the most powerful proofs of AI commercial value. Lo Toney commented on this: "The key is who can win the market for office assistants. If the larger market develops towards AI agents and their coordination - reasoning infrastructure, multimodal workflows, enterprise search - then Alphabet Inc. Class C will have a huge opportunity to drive Alphabet's future growth." Intelligent Shopping: From search engines to closed-loop transactions The e-commerce sector may be the most disruptive focus of this year's I/O. In January 2026, Alphabet Inc. Class C officially announced at the NRF Big Show its entry into the era of "Agentic Commerce," launching the Universal Commerce Protocol (UCP), a new open standard aimed at becoming a "universal language" for collaborative operations between intelligent agents and systems on consumer interfaces, enterprises, and payment providers. The protocol was developed with the participation of major retailers such as Shopify, Target, Walmart Inc., and Wayfair, and is supported by payment networks such as MasterCard, American Express Company, and Visa. In February of this year, Alphabet Inc. Class C further embedded AI shopping functions into its search engine and Gemini chatbot, allowing users to directly purchase products in AI-driven answers, achieving a one-stop shopping experience called "ask-and-buy." Brands can also offer exclusive discounts to potential customers through the "direct discount" feature. Sameer Samat, Head of the Alphabet Inc. Class C Android ecosystem, described this scenario by letting Gemini help him plan a barbecue, create a menu, open Instacart, add ingredients to the Safeway shopping cart, and notify him when the task is completed - "If you have to perform these operations multiple times every day for a week, you can save a lot of time." For investors, the impact of this strategy goes far beyond Alphabet itself - Mizuho Securities pointed out that Alphabet Inc. Class C developing more intelligent products may put pressure on e-commerce platforms such as Booking Holdings, Expedia, DoorDash, Zillow, and Instacart, and expectations for this shift may already be reflected in recent weakness in these stocks. AI Search Monetization: A battle over "click-through rates" Search advertising is the core cash cow for Alphabet Inc. Class C, and the introduction of AI models is profoundly changing the underlying logic of this business. In the first quarter of 2026, Alphabet Inc. Class C's search advertising revenue reached $60.4 billion, a 19% year-over-year increase, with query volume reaching a historical high. Citi data shows that AI-driven ad campaigns currently account for over 30% of search ad spend; the AI Max tool, which ended its beta testing in April, is planned to replace dynamic search ads in September with a full set of features that have resulted in higher conversion rates. However, the flip side of the coin is also worth noting. Mizuho Securities points out that AI-modeled search has led to a significant decrease in external click-through rates - the company estimates that 93% of searches ultimately did not generate external clicks, and the natural click-through rate of AI overview queries has also decreased by 15%. This means that Alphabet Inc. Class C is transitioning from a "traffic distributor" to an "attention capturer," and how to maintain the value perception of advertisers in this transition will be one of the most closely watched issues at I/O and the subsequent Google Marketing Live conference on May 21. Deepwater Asset Management's Managing Partner Gene Munster has stated that he will focus on new advertising products in AI mode, how Alphabet Inc. Class C is building intelligent commerce, and what insights it provides for a more personalized AI experience. Alphabet Inc. Class C Cloud: Leading with a 63% growth rate, backlog orders nearly double For investors, the most influential I/O news may come from the cloud computing and infrastructure sector. Alphabet Inc. Class C Cloud has become the strongest growth engine for Alphabet - in the first quarter of 2026, Alphabet Inc. Class C Cloud achieved revenue of $20.28 billion, a 63% year-over-year increase, significantly exceeding market expectations of around $18 billion. This growth rate not only surpasses the 48% growth rate in the previous quarter but also significantly outperforms Microsoft Corporation's Azure (about 30%) and AWS (28%). What is even more exciting for Wall Street is that Alphabet Inc. Class C Cloud's backlog orders almost doubled quarter-over-quarter in the first quarter, reaching $462 billion, with about half expected to confirm revenue within the next 24 months. Product revenue built on generative AI has grown nearly 800% year-over-year. Sundar Pichai noted in the earnings call that the number of deals worth over a billion dollars signed in 2025 exceeded the total of the past three years, with current customers purchasing more than 30% above the promised amount. In terms of profit margins, Alphabet Inc. Class C Cloud's operating profit margin has reached 32.9%, an increase of 15.7 percentage points year-over-year, proving its ability to continue expanding profit margins while making large-scale investments. AI Chips: TPU external sales open up a whole new revenue curve Perhaps the most underestimated variable at this year's I/O is the external sales of Alphabet Inc. Class C's TPU chips. In the first quarter of this year, Alphabet Inc. Class C first disclosed that it had delivered TPUs to a group of customers in hardware configuration form, with related agreements reflected in a backlog order of $462 billion. CFO Annette Ashkenazi stated that the company would begin recognizing a small portion of TPU sales revenue in the second half of 2026, with most sales expected to be realized in 2027. At the Cloud Next conference in April, Alphabet Inc. Class C officially launched the eighth-generation TPU, split into the TPU 8t for training and the TPU 8i for inference. The TPU 8t cluster can be scaled up to 9,600 chips, equipped with 2PB of shared high-bandwidth memory, achieving a nearly three-fold increase in computing performance compared to the previous generation; the TPU 8i is equipped with 288GB of HBM, achieving an 80% cost-effectiveness improvement through its "memory wall-breaking" design. Lo Toney believes that TPUs are one of the most underestimated parts of Alphabet's vision, and that Alphabet Inc. Class C's proprietary chips enable the company to build a tightly integrated AI infrastructure. Gene Munster estimates that the broader AI chip market is worth around $500 billion annually, meaning that even moderate market share growth could have a significant impact on Alphabet. Anthropic Relationship: A "double hedge" of $200 billion In the run-up to I/O, Alphabet Inc. Class C's relationship with Anthropic has become the focus of the market. On May 6, according to The Information, Anthropic has committed to paying about $200 billion to Alphabet Inc. Class C Cloud over the next five years to purchase cloud services and TPU computational power. This transaction accounts for over 40% of the $462 billion backlog of orders at Alphabet Inc. Class C Cloud. This transaction's structure is quite ingenious: Alphabet Inc. Class C initially invested $10 billion in cash (based on a valuation of $350 billion) into Anthropic, with an additional $40 billion if certain milestones are met; meanwhile, Anthropic has committed to spending $200 billion on cloud services and TPU purchase. Funds flow from Alphabet Inc. Class C to Anthropic, and then flow back to Alphabet Inc. Class C through cloud services and TPU expenses, creating a unique "fund cycle." Lo Toney described this relationship more like a "hedge" - even if companies are more inclined to use Claude rather than Gemini, Alphabet Inc. Class C can still benefit from infrastructure needs: "If companies are more inclined to use Claude, Alphabet Inc. Class C will still come out on top in terms of infrastructure because all of these activities still need to be hosted somewhere. Alphabet Inc. Class C still wins because they have TPUs." Conclusion: From "followers" to "definers"? The significance of this year's Alphabet Inc. Class C I/O conference lies not only in being a product launch event but also in testing the pressure of Alphabet's "full-stack AI" strategy. Wall Street has given Alphabet a generous valuation premium - with the stock price rising 140% in a year, continuously setting new 52-week highs - but all of this is conditional on AI truly transforming into sustained revenue growth for search, advertising, shopping, and cloud services. Munster succinctly captured the essence of this moment: "Having a complete technology stack gives advantages in innovation speed. When you develop with your own custom chips, the speed advantage comes into play. When you have ample power resources, you can build data centers faster. Speed advantage is crucial."