Alphabet Crosses $400B in Revenue, Ships Gemini 3.1 Pro, and Doubles Down on a $185B AI Capex Bet

Alphabet just posted its first $400B revenue year, with Google Cloud accelerating to 48% growth and Gemini models processing 10 billion tokens per minute. But the real story is the $175-185B capex guidance for 2026 — nearly double last year — as the company races to own the full AI stack from TPUs to frontier models to consumer products.

GOOGL · Communication Services · February 19, 2026

S&P 500 Position

Dominates the Communication Services sector alongside Meta Platforms. Within the S&P 500's top 10, Alphabet displaced Apple for the #2 slot as of January 2026, reflecting the market's repricing of AI infrastructure plays. The competitive dynamic with Meta is direct in advertising (Google's $260B+ vs Meta's comparable ad revenue) but divergent in platform strategy: Google owns the AI stack vertically (chips to models to products) while Meta bets horizontally on open-source models and social graph. Netflix and Walt Disney are sector neighbors but operate in fundamentally different businesses.

Index Weight: ~6.39% | Rank: #2 in S&P 500 (combined Class A + C shares); behind only NVIDIA

Company Overview

Alphabet is executing the most aggressive vertical integration play in AI history. The company controls every layer of the stack: custom silicon (TPUs since 2014), frontier foundation models (Gemini 3.x family), developer platforms (Antigravity, Vertex AI), consumer surfaces (Search, YouTube, Android), and distribution at planetary scale. In Q4 2025, Gemini models processed over 10 billion tokens per minute via direct API use — up from 7 billion the prior quarter — while the Gemini App reached 750 million monthly active users. Gemini 3.1 Pro launched today (February 19, 2026), scoring 77.1% on ARC-AGI-2 and outperforming Anthropic's Opus 4.6 and OpenAI's GPT-5.2 on the majority of cited benchmarks. The financial engine remains advertising, but the growth story has shifted to Cloud and AI. Google Cloud revenue accelerated to 48% year-over-year growth in Q4, reaching a $70B+ annualized run rate with a backlog that more than doubled to $240 billion. YouTube surpassed $60 billion in combined ad and subscription revenue for FY2025. Meanwhile, Waymo completed 15 million trips across six U.S. cities, and Isomorphic Labs (led by Nobel laureate Demis Hassabis) is entering Phase 1 clinical trials for AI-designed cancer drugs. The DOJ's antitrust case resulted in behavioral remedies rather than a forced Chrome/Android divestiture, but the government has cross-appealed, keeping regulatory uncertainty alive through 2026-2027.

Products & Revenue

Alphabet's revenue machine has three segments but one dominant engine. Google Services (Search ads, YouTube ads, subscriptions, platforms, and devices) generates roughly 84% of total revenue, powered by the world's largest advertising ecosystem. Google Cloud — spanning GCP infrastructure, Vertex AI, Gemini Enterprise, and Google Workspace — now contributes about 15% and is the fastest-growing segment, driven by enterprise AI demand. Other Bets (primarily Waymo and Verily) accounts for less than 1% of revenue but consumes substantial R&D investment.

Google Search & Other Advertising (56%): Core search advertising plus Shopping, Maps, and AI Mode monetization. Q4 2025 revenue of $63.1B was up 17% YoY, driven by broad vertical strength — particularly retail — and expanding AI Overviews.

YouTube Advertising (10%): Video ad revenue across in-stream, Shorts, and direct response formats. Shorts now averages over 200 billion daily views and earns more revenue per watch hour than traditional in-stream in multiple markets including the U.S.

Google Network Advertising (7%): Ad revenue from partner properties via AdSense and Ad Manager. Declining modestly at -2% YoY as the industry shifts spend toward owned-and-operated surfaces.

Google Subscriptions, Platforms & Devices (12%): YouTube Premium/Music/TV, Google One, Google Play transactions, and Pixel hardware. Over 325 million paid subscriptions across consumer services. Growing 17% YoY, fueled by Google One AI plan demand.

Google Cloud (15%): GCP compute/storage/networking, Vertex AI platform, Gemini Enterprise, Google Workspace. 48% YoY growth in Q4 2025 with 14 product lines each exceeding $1B in annual revenue and operating margins expanding to 30.1%.

Other Bets (<1%): Primarily Waymo (autonomous ride-hailing) and Verily (life sciences). Waymo served 15 million trips in 2025 across six U.S. cities and raised $16B in February 2026. Revenue is minimal; operating loss was $3.6B in Q4 including a $2.1B Waymo SBC charge.

Based on Alphabet FY2025 10-K and Q4 2025 earnings release dated February 4, 2026. Revenue percentages approximated from full-year $403B consolidated revenue.

Leadership

Sundar Pichai

CEO since 2019 (Alphabet CEO); 2015 (Google CEO). Pichai joined Google in 2004 and led Chrome, Chrome OS, and Google Drive before ascending to CEO. He holds a metallurgical engineering degree from IIT Kharagpur, an M.S. from Stanford, and an MBA from Wharton. He has steered Alphabet's transformation into an AI-first company, overseeing the merger of Google Brain and DeepMind and the launch of the Gemini model family.

Demis Hassabis, CEO & Co-Founder, Google DeepMind; CEO, Isomorphic Labs: Nobel Prize in Chemistry laureate (2024) for AlphaFold's protein structure prediction. Leads all of Alphabet's frontier AI research and the Gemini model family. Isomorphic Labs is entering Phase 1 clinical trials for AI-designed cancer drugs with a pipeline of 17 active drug programs and partnerships with Eli Lilly and Novartis worth ~$3B in milestone payments.

Thomas Kurian, CEO, Google Cloud: Former Oracle President of Product Development who joined in 2019 when Cloud was a distant #3. Has transformed Google Cloud into a $70B+ run rate business with 48% growth, 14 billion-dollar product lines, and a $240B backlog. Princeton EE and Stanford MBA with 22 years at Oracle prior.

Anat Ashkenazi, CFO, Alphabet: Joined in 2024 from Eli Lilly where she served as EVP and CFO. Overseeing the most capital-intensive period in Alphabet's history: $91.4B capex in 2025 with $175-185B guided for 2026. Managed a $24.8B bond issuance in November 2025 and a subsequent $20B raise in February 2026.

Philipp Schindler, SVP & Chief Business Officer, Google: Oversees Google's global advertising sales and operations — the engine behind ~75% of all Alphabet revenue. Leading the integration of AI into ad products including AI Max, Performance Max, and the monetization of AI Mode in Search.

Ruth Porat, President & Chief Investment Officer, Alphabet: Former CFO (2015-2024), now architecting Alphabet's long-term capital deployment strategy. Managing the $20B February 2026 bond issuance — which included a landmark 100-year £1B 'Century Bond' — while maintaining the company's AA+ credit rating.

The AI Angle

Full-stack AI dominance from silicon to consumer

Alphabet owns the most vertically integrated AI stack in the industry. It starts at the silicon layer with custom Tensor Processing Units (TPUs, in development since 2014) and extends through foundation models (Gemini 3.x family), developer platforms (Vertex AI, Antigravity, Gemini API), and consumer products with planetary distribution (Search, YouTube, Android, Chrome). The Gemini 3 Pro model — launched in late 2025 — demonstrated the fastest adoption of any model in company history, consistently processing three times as many daily tokens as its predecessor 2.5 Pro. Gemini 3.1 Pro shipped today (February 19, 2026) with a massive leap on ARC-AGI-2 from 31.1% (3 Pro) to 77.1%, and outperforms rival commercial models from Anthropic and OpenAI on the majority of benchmarks. The company reduced Gemini serving unit costs by 78% over 2025 through model optimizations and infrastructure efficiency. Nearly 350 customers each processed more than 100 billion tokens in December alone. The infrastructure commitment is staggering. Alphabet spent $91.4B in capex during 2025 (60% servers, 40% data centers and networking) and guided $175-185B for 2026 — nearly doubling year-over-year. This is funded by operating cash flow ($52.4B in Q4 alone, a company record) and aggressive debt issuance including $24.8B in November 2025 and $20B in February 2026 (the latter featuring a 100-year Century Bond). Google Cloud's AI infrastructure business is the primary beneficiary: the segment's backlog more than doubled year-over-year to $240B, and the number of billion-dollar deals in 2025 surpassed the previous three years combined. Google Cloud now supports both first-party models (Gemini) and third-party models on Vertex AI, giving enterprises model flexibility — a strategic advantage over more opinionated competitors. The research arm, Google DeepMind under Demis Hassabis, operates as the most prolific AI research organization in the world. Gemini Deep Think achieved gold-medal level performance across the International Math, Physics, and Chemistry Olympiads. More practically, DeepMind is now applying these capabilities to professional research: autonomously solving open mathematical conjectures, assisting in reviewing CS theory papers for STOC'26, and — through Isomorphic Labs — entering Phase 1 clinical trials for AI-designed cancer drugs with a 17-program pipeline. Hassabis, speaking today at the India AI Impact Summit, predicted AGI within five to eight years. Google Antigravity, the company's new agentic development platform, surpassed 1.5 million weekly active users within two months of launch. The Gemini Robotics-ER 1.5 model previewed in late 2025 signals the next frontier: physical intelligence. The competitive risks are real. OpenAI's GPT-5.x family and Anthropic's Opus 4.6 remain formidable on specific benchmarks (SWE-Bench, Humanity's Last Exam). DeepSeek's V4 model — released in early February 2026 — demonstrated frontier-level capability at a fraction of Western costs, triggering investor anxiety about capex ROI across the sector. The DOJ's cross-appeal of the antitrust remedy keeps alive the possibility of more severe structural interventions. And the fundamental question of search cannibalization persists: if AI-generated answers replace ad-laden search results pages, Alphabet's highest-margin business faces long-term pressure. The company is experimenting with AI Mode monetization — testing ads below AI responses and direct offers in Google Ads — but the transition from click-based to AI-based monetization remains the defining business model risk of this era.

Financial Snapshot

Revenue (TTM): $402.9B — TTM ending December 31, 2025 | Net Income: $132.2B net income

Margins: Operating 31.6% (Q4 2025), net 32.8% (TTM)

Alphabet crossed $400B in annual revenue for the first time in FY2025, growing 15% on a constant-currency basis. The margin story is nuanced: Google Cloud margins expanded dramatically (30.1% operating margin in Q4, up from 17.5% a year ago), partially offset by elevated shared R&D costs for Gemini development and the $2.1B Waymo SBC charge. Free cash flow of $24.6B in Q4 alone demonstrates the business's cash generation capacity even amid massive infrastructure investment. The $175-185B capex guide for 2026 is the key investor debate: bulls see it as securing decade-long infrastructure moats; bears see potential margin compression if AI revenue growth decelerates.

1-Year Performance

GOOGL at $303.33, up approximately 64% over the past year. The stock hit an all-time closing high of $343.69 on February 2, 2026, then pulled back ~12% amid the broader 'Great Sector Rotation' away from mega-cap tech.

The stock's 64% one-year gain was fueled by accelerating Cloud revenue, the favorable antitrust ruling in September 2025 (stock jumped 8% after Chrome divestiture was rejected), and the Q4 beat. The recent pullback from all-time highs reflects two forces: the DeepSeek V4 shock in early February 2026, which triggered broad capex-fatigue selling across AI infrastructure names, and a sector rotation into cyclical 'Old Economy' stocks. The DOJ's cross-appeal filed on February 4, 2026 added regulatory uncertainty. Analyst consensus remains bullish with a median price target of ~$373.

Fun Fact: Alphabet's custom TPU program — which now underpins its entire Gemini inference and training stack — began in 2014 under the internal codename 'ASIC,' years before the current AI boom. Google Cloud CEO Thomas Kurian has noted that his identical twin brother, George Kurian, is CEO of NetApp — making them the only known pair of identical twin siblings simultaneously running major publicly-traded technology companies.