Alphabet 营收增长 24%，得益于 AI 投资，搜索增长 17%，云增长 82%，Gemini 模型快速普及。
Alphabet reports 24% revenue growth driven by AI investments, with Search up 17%, Cloud up 82%, and Gemini models seeing rapid adoption.
要点 · TL;DR
AI 投资推动 Alphabet 2026 年第二季度营收增长 24%. AI investments drove 24% revenue growth across Alphabet in Q2 2026.
Google Cloud 营收增长 82%，积压订单达 5140 亿美元，受企业 AI 需求推动. Google Cloud revenue surged 82% with $514B backlog, fueled by enterprise AI.
Gemini 模型深度集成到搜索、云和 YouTube 中，提升了变现能力. Gemini models are deeply integrated into Search, Cloud, and YouTube, boosting monetization.
核心观点 · Key points
AI 投资正在重新定义 Alphabet 业务的每个部分，推动收入增长 24%。 AI investments are redefining every part of Alphabet's business, driving 24% revenue growth.
Google Cloud 收入增长 82%，积压订单达 5140 亿美元，受企业 AI 需求驱动。 Google Cloud revenue grew 82% with backlog reaching $514 billion, driven by enterprise AI demand.
Gemini 模型深度集成到搜索、云和 YouTube 中，改善了变现和用户体验。 Gemini models are deeply integrated across Search, Cloud, and YouTube, improving monetization and user experience.
由于供应限制和强劲的 AI 需求，2026 年资本支出指引上调至 1950-2050 亿美元。 CapEx guidance raised to $195-205 billion for 2026 due to supply constraints and strong AI demand.
Alphabet 仍致力于前沿模型开发，Gemini 4 的预训练显示出有希望的进展。 Alphabet remains committed to frontier model development with Gemini 4 pre-training showing promising progress.
反共识 · Contrarian takes
Alphabet 使用第三方容量作为过渡策略，尽管面临利润率压力，但优先考虑长期客户价值。 Alphabet uses third-party capacity as a bridging strategy despite margin pressure, prioritizing long-term customer value.
向客户数据中心销售 TPU 系统扩大了总可寻址市场，但收入确认集中在 2027 年。 TPU system sales to customer data centers expand TAM but revenue recognition is back-end loaded to 2027.
Alphabet 通过股权和债务融资以维持弹性资产负债表，而不仅仅是为了增长资金。 Alphabet raised equity and debt to maintain a resilient balance sheet, not just for growth funding.
Gemini Flash 模型优先考虑性价比，而前沿领导力需要更大的基础模型，如 Gemini 4。 Gemini Flash models are prioritized for cost-performance, while frontier leadership requires larger base models like Gemini 4.
Alphabet 首先将算力分配给 AGI 开发，然后是核心产品，平衡内部和外部需求。 Alphabet allocates compute first to AGI development, then to core products, balancing internal and external demand.