[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-gemini-app-1b-monthly-users":3,"news-related-8b6c20ec-7222-48cf-af2c-ac97466a2b0a":38},{"id":4,"title":5,"summary":6,"content":7,"original_url":8,"source_id":9,"tags":10,"translations":24,"news_slug":31,"published_at":32,"created_at":33,"modified_at":34,"is_published":35,"publish_type":36,"image_url":14,"view_count":37},"8b6c20ec-7222-48cf-af2c-ac97466a2b0a","Gemini 月活破 10 亿:Google 第一次把 AI 助手做成自家「最快十亿用户产品」","Google CEO 桑达尔·皮查伊 8 月 11 日宣布 Gemini App 月活用户突破 10 亿,成为 Google 历史上最快跨过这一门槛的产品,也是首个围绕自研前沿模型打造的十亿用户产品。Google 当日同步披露使用数据:63% 用户通过语音与 Gemini 对话,1\u002F5 的 Gemini Live 会话带摄像头或屏幕共享,每日生成 1.5 亿张图像,iOS 端月活也超过 1 亿。","# Gemini 月活破 10 亿:Google 把 AI 助手做成了「最快十亿用户产品」\n\nGoogle CEO 桑达尔·皮查伊(Sundar Pichai)8 月 11 日在 X 上宣布,**Gemini App 的月活用户(MAU)正式突破 10 亿**——这是 Google 历史上最快达成 10 亿用户的产品,也是 Google 第一个围绕自研前沿模型打造的「十亿用户级」产品([blog.google](https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F))。皮查伊同时点名负责 Google Labs、Gemini App 与 AI Studio 的副总裁 Josh Woodward,以及整个 Gemini 团队,称这是「分布最广的 AI 助手」的一次标志性时刻([unite.ai](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F))。\n\n## 数据要点\n\nGoogle 在 Keyword 官方博客同步披露了一组使用数据,勾勒出一个「已经走出文本框」的产品:\n\n- **63% 的用户直接用语音与 Gemini 对话**,其中「纯语音用户」比例在持续上升,繁忙的父母使用语音的比例高出普通用户 43%。\n- **1\u002F5 的 Gemini Live 会话超出了语音**,会带入摄像头画面或屏幕共享,做实时问题求解(Google 称这在 DIY 圈和学生群体里特别受欢迎)。\n- **38% 的学生请求附带附件上传**,Google 计划未来几周推出一批面向学习的工具。\n- **每天生成超过 1.5 亿张图像**;图像、视频、音频可在同一界面完成,被小企业用来做营销物料。\n- **iOS 端月活超过 1 亿**,macOS 高频用户在 Gemini 上的提问频次大约是其他端用户的两倍。\n- **Android 端可串联 40+ 主流 App 自动化**,覆盖打车、订餐等高频动作([blog.google](https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F))。\n\n按 Google 自家披露的曲线推算,Gemini App 在 2025 年 5 月 I\u002FO 大会时月活约 4 亿,2025 年 Q3 财报(2025 年 10 月 29 日)已超过 6.5 亿,2026 年 5 月 I\u002FO 大会时突破 9 亿,7 月 22 日 Q2 财报披露月活 9.5 亿;**距离 8 月 11 日公布 10 亿,只过了三周**——最后冲刺的 5000 万用户在 21 天内完成。从 I\u002FO 2025 到今天,15 个月内净增约 6 亿月活([unite.ai](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F))。\n\n## 增长背后是「Voice + Agent」的双引擎\n\nGoogle 把这次的 10 亿归因到两类能力:**Voice**(Gemini Live 的多模态对话)和 **Agentic**(Gemini Spark 这类常驻 Agent)。\n\n- Gemini Live 在 I\u002FO 2026 上重做,Google 描述它「更聪明、更快、对背景噪声更不敏感」,这次公布的 63% 语音占比与 20% Live 摄像头\u002F屏幕共享会话,都直接挂在这条产品线上。\n- Gemini Spark——常驻型个人 Agent,跑在 Gemini 3.5 与 Antigravity harness 上——5 月 I\u002FO 上以受信任测试者形式上线,7 月 31 日「Gemini Drop」开始全球推送(欧洲经济区、英国、瑞士、尼日利亚除外),同期还发布了 Gemini 3.6 Flash 与 Gemini 3.5 Flash-Lite。\n\n需要注意一些细节:Spark 的全球推送**排除了欧洲经济区、英国、瑞士和尼日利亚**;搜索里的 Information Agent 先给 Google AI Pro 与 Ultra 订户;Daily Brief 进入 Gemini App 是 I\u002FO 上预告的内容。这两条信息说明「10 亿」背后仍有一段监管和分层发布的尾巴。\n\n## 「14 个十亿产品」俱乐部和那个值得记住的注脚\n\nPichai 在 I\u002FO 2026 上曾说 Google 有 13 个 10 亿用户级产品,其中 5 个超过 30 亿。Gemini 是第 14 个,也是**第一个原生围绕 Google 自研前沿模型打造的十亿用户产品**。从分发角度,这意味着在所有 AI 实验室里,Google 现在拥有最完整的「自研模型 + 自有 10 亿分发」组合([unite.ai](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F))。\n\n但 Ars Technica 同日发的稿件给这次里程碑加了一个「冷却」注脚:Gemini 在 6 月原定发布 Gemini 3.5 Pro 的窗口已经错过,Google 7 月承认 3.5 Pro 仍在测试,并已开始训练 Gemini 4;DeepMind 联合创始人 Demis Hassabis 已从 CEO 位退到首席科学家和董事长;近期多名核心研究员和科学家离职。**「10 亿 MAU 是真,但 Gemini 在前沿模型节奏上确实在变慢」**——这条裂缝如果继续扩大,迟早会反映在下个季度的留存和活跃度上([arstechnica.com](https:\u002F\u002Farstechnica.com\u002Fai\u002F2026\u002F08\u002Fgoogle-says-gemini-has-reached-1b-users-faster-than-any-other-google-product\u002F))。\n\n## 我们的观察\n\n这次「Gemini 10 亿月活」值得拆成三层来看:\n\n1. **分发层面**:Google 靠 Android 出厂预装、跨应用 Agent、Web 集成(Chrome、Search 的 AI Mode \u002F Overviews)拿到的不只是「10 亿用户」,而是一个**持续在线、可观察、可升级**的分发通道——这是任何闭源 LLM 厂商都买不到的存量。\n2. **产品形态**:10 亿人里 63% 在用语音,1\u002F5 在共享摄像头\u002F屏幕,1.5 亿张图\u002F天——这意味着 Gemini 已经不只是「问问题」的工具,正在变成**嵌入生活节奏的常驻 Agent**(预订、归纳、生成、做家务、帮学生写作业)。\n3. **潜在风险**:分发优势是「存量」,模型能力是「增量」。Hassabis 退任、Gemini 3.5 Pro 难产、研究员流失这些信号叠加在一起,提醒我们:**当一个 AI 助手的「好用程度」被 ChatGPT、Claude、Kimi、Qwen 拉开时,10 亿 MAU 也会松动**——下个季度的 DAU\u002FMAU 比例,会比 MAU 本身更能说明 Google 的真实处境。\n\n所以,10 亿不是终点,而是「自研模型 vs 分发优势」这条战线的新起点。\n\n参考来源:\n- [Google Keyword 官方博客《More than 1 billion people are using the Gemini app every month》(2026-08-11)](https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F)\n- [Unite.AI: Gemini App Tops 1 Billion Monthly Users as Google's Fastest-Growing Product (2026-08-11)](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F)\n- [Ars Technica: Gemini becomes Google's fastest-growing product ever as it hits 1B users (2026-08-11)](https:\u002F\u002Farstechnica.com\u002Fai\u002F2026\u002F08\u002Fgoogle-says-gemini-has-reached-1b-users-faster-than-any-other-google-product\u002F)","https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F","4d11edad-2df6-45f6-b71f-70f65de7f7fd",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"5e628969-6d2a-437f-998a-104e4b16cfb1","ai-progress",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"a9524a82-a7c5-4daa-bb4b-a7ee77bb0b94","gemini",{"id":19,"name":20,"slug":20,"description":14,"color":14},"8cf7490f-2449-4ba7-be19-61befa0d92b4","google",{"id":22,"name":23,"slug":23,"description":14,"color":14},"01598627-1ea6-4b27-a5d8-874971571a71","llm",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"9a26c25e-f01b-41d3-b907-7339630a3b24","en","Gemini passes one billion MAU, Google's fastest product yet","Google CEO Sundar Pichai announced on August 11 that the Gemini App has passed 1 billion monthly active users, making it the fastest product in Google's history to reach that milestone and the company's first billion-user product built natively around its own frontier models. Google also disclosed that 63% of users talk to Gemini directly, 1 in 5 Gemini Live sessions involve camera or screen-share, 150M+ images are generated per day, and iOS-side MAU has surpassed 100M.","# Gemini App crosses 1B MAU: Google turns its AI assistant into the fastest product to a billion\n\nGoogle CEO Sundar Pichai announced on August 11 that **the Gemini App has officially passed 1 billion monthly active users (MAU)** — making it the fastest product in Google's history to cross that threshold, and the company's first billion-user product built natively around its own frontier models ([blog.google](https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F)). Pichai named Josh Woodward, the VP leading Google Labs, the Gemini app, and AI Studio, alongside the broader Gemini team, calling it a defining moment for \"the most widely distributed AI assistant\" ([unite.ai](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F)).\n\n## What the data shows\n\nGoogle posted a companion set of usage numbers on its Keyword blog that sketch a product long past the text-box phase:\n\n- **63% of users now talk to Gemini directly**, with rising \"voice-only\" usage; busy parents are 43% more likely to use voice for everyday tasks.\n- **1 in 5 Gemini Live sessions go beyond voice**, pulling in live camera feeds or screen sharing for real-time problem-solving — Google says DIYers and students use this heavily.\n- **38% of student requests include an attachment upload**, and Google is preparing a new batch of study tools in the coming weeks.\n- **Over 150 million images generated per day**; images, video, and audio live in one interface, and small businesses use it for marketing assets.\n- **More than 100 million MAUs on iOS**, and macOS power users prompt about 2x as often as users on other surfaces.\n- **Android-side automation spans 40+ popular apps** — ride-hailing, restaurant reservations, and more ([blog.google](https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F)).\n\nBy Google's own numbers: ~400M MAU at I\u002FO 2025 (May 2025); 650M+ at Q3 2025 earnings (Oct 29, 2025); 900M+ at I\u002FO 2026 (May 19, 2026); 950M in the Q2 2026 release filed July 22, 2026; **1B on Aug 11 — a 50M jump in 21 days and a 600M jump in 15 months** ([unite.ai](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F)).\n\n## The \"Voice + Agent\" twin engines\n\nGoogle attributes the milestone to two product capabilities: **Voice** (Gemini Live's multimodal conversation) and **Agentic** (always-on agents like Gemini Spark).\n\n- **Gemini Live** was rebuilt at I\u002FO 2026 — Google describes it as \"smarter, faster, and less distracted by background noise.\" The 63% voice share and 20% camera\u002Fscreen-share sessions hang directly off this line.\n- **Gemini Spark**, the always-on personal agent running on Gemini 3.5 with the Antigravity harness, moved from a trusted-tester launch at I\u002FO to a worldwide rollout in the July 31 \"Gemini Drop,\" which also shipped Gemini 3.6 Flash and Gemini 3.5 Flash-Lite.\n\nTwo caveats: Spark's global rollout **excludes the European Economic Area, the UK, Switzerland, and Nigeria**; Information Agents in Search launch first to Google AI Pro and Ultra subscribers; Daily Brief in the Gemini app is on the I\u002FO roadmap. The \"1 billion\" headline sits on top of a regulated, tiered rollout.\n\n## The 14-member billion-user club and the footnote that matters\n\nPichai said at I\u002FO 2026 that Google had 13 products above a billion users, with 5 above 3 billion. Gemini is the 14th — and the **first natively built around Google's own frontier models**. In distribution terms, Google now owns the only complete stack of \"self-trained model + proprietary billion-user surface\" among AI labs ([unite.ai](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F)).\n\nBut Ars Technica's same-day piece added a cooling footnote: Gemini's June window for releasing Gemini 3.5 Pro has already slipped; in July Google admitted 3.5 Pro is still in testing while Gemini 4 training has begun; DeepMind co-founder Demis Hassabis has stepped down from CEO to Chief Scientist \u002F Chairman; multiple senior scientists and researchers have left recently. **\"1B MAU is real, but the frontier model cadence is genuinely slowing.\"** If that gap widens, it will eventually show up in next quarter's retention and DAU\u002FMAU ([arstechnica.com](https:\u002F\u002Farstechnica.com\u002Fai\u002F2026\u002F08\u002Fgoogle-says-gemini-has-reached-1b-users-faster-than-any-other-google-product\u002F)).\n\n## What to take away\n\nThree layers to read this milestone through:\n\n1. **Distribution**: Google's Android pre-install footprint, cross-app agent hooks, and Web integrations (Chrome, Search AI Mode \u002F Overviews) didn't just buy \"1B users\" — they bought a **continuously online, observable, upgradable distribution channel** that no closed-source LLM competitor can purchase.\n2. **Product shape**: 63% voice, 1 in 5 live camera\u002Fscreen-share sessions, 150M images\u002Fday — Gemini has moved past \"ask a question\" toward an **ambient agent woven into daily routines** (bookings, summarization, generation, homework, chores).\n3. **Latent risk**: distribution is \"stock\"; model capability is \"flow.\" Hassabis stepping aside, 3.5 Pro slipping, and researcher attrition together remind us that **when an assistant's quality falls behind ChatGPT, Claude, Kimi, or Qwen, even 1B MAU gets soft** — next quarter's DAU\u002FMAU ratio will tell more about Google's real position than MAU itself.\n\nSo 1 billion is not the finish line. It is the new starting gun for the \"frontier model vs. distribution advantage\" race.\n\nSources:\n- [Google Keyword blog — \"More than 1 billion people are using the Gemini app every month\" (2026-08-11)](https:\u002F\u002Fblog.google\u002Finnovation-and-ai\u002Fproducts\u002Fgemini-app\u002Fone-billion-monthly-users\u002F)\n- [Unite.AI — \"Gemini App Tops 1 Billion Monthly Users as Google's Fastest-Growing Product\" (2026-08-11)](https:\u002F\u002Fwww.unite.ai\u002Fgemini-app-tops-1-billion-monthly-users-as-googles-fastest-growing-product\u002F)\n- [Ars Technica — \"Gemini becomes Google's fastest-growing product ever as it hits 1B users\" (2026-08-11)](https:\u002F\u002Farstechnica.com\u002Fai\u002F2026\u002F08\u002Fgoogle-says-gemini-has-reached-1b-users-faster-than-any-other-google-product\u002F)","gemini-app-1b-monthly-users","2026-08-12T03:00:00Z","2026-08-12T14:05:25.948158Z","2026-08-12T14:05:25.948168Z",true,"agent",84,{"items":39},[40,45,50,55,60,65],{"id":41,"title":42,"news_slug":43,"published_at":44},"7b9cdf6e-5ef0-4ece-ab6c-e8cec1b02397","Google 重组 DeepMind 领导层,Gemini 研发提速应对 Anthropic 与 OpenAI 竞争","google-deepmind-reshuffle-gemini-speed","2026-08-25T07:00:00+00:00",{"id":46,"title":47,"news_slug":48,"published_at":49},"bcedeb8e-e5eb-4bbc-98b8-ea12f869055f","Google 收编 DeepMind：25 年最大 AI 重组","google-deepmind-centralization-gemini-catchup","2026-08-14T08:00:00+00:00",{"id":51,"title":52,"news_slug":53,"published_at":54},"4bd8e8bd-7066-4ab7-bd97-e24ea3921395","Gemini 因编程落后推迟两月:Brin 4 月督促背后,Google 把研发「收回到一个人」手里的组织账本","google-gemini-coding-behind-deepmind-reshuffle","2026-08-14T03:30:00+00:00",{"id":56,"title":57,"news_slug":58,"published_at":59},"34edaffc-6b5c-4df1-9e2f-d864cada6063","Gemini 走进 K-12 课堂：Google 把「上下文」塞进每个作业","gemini-classroom-k12-contextualized-prompts","2026-08-07T02:00:00+00:00",{"id":61,"title":62,"news_slug":63,"published_at":64},"df25ef84-5f0b-477b-9940-a0da477d8169","DeepMind 新主帅接棒：Hassabis 退任，Gemini 4 成 Google 筹码","deepmind-kavukcuoglu-gemini-4-reshuffle","2026-08-07T00:00:00+00:00",{"id":66,"title":67,"news_slug":68,"published_at":69},"d8a15300-7bba-40d9-a62c-436a2f3a68bc","Google ATLAS 报告:1500 万次 Gemini 交互告诉我们什么?","google-gemini-atlas-work-adoption-data","2026-07-30T03:30:00+00:00"]