<p>根據市場情報公司 Sensor Tower 發佈的年度《移動狀態》報告，2025 年消費者在非遊戲移動應用上的支出首次超過了遊戲支出。雖然這一里程碑在某些市場（如美國）或特定季度中曾出現，但 2025 年標誌着這一現象在全球範圍內的首次發生。全球消費者去年在應用上的支出約為 850 億美元，同比增長 21%。這一數字也幾乎是五年前支出的 2.8 倍。</p>
<p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.01-PM.jpg/query-dz02ODA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.01-PM.jpg/query-dz02ODA"/></p>
<p><strong>圖片來源：</strong> Sensor Tower</p>
<p>生成性人工智能作為過去一年中的一個重要趨勢，推動了收入增長，因為這一類別的應用內購買收入在 2025 年超過了 50 億美元，增長了三倍多。AI 應用的下載量也在增長，年同比翻了一番，達到了 38 億次。</p>
<p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.33.06-PM.jpg/query-dz02ODA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.33.06-PM.jpg/query-dz02ODA"/></p>
<p><strong>圖片來源：</strong> Sensor Tower</p>
<p>該領域的增長可以歸因於多個因素。首先，消費者對 AI 助手的熱情是一個重要驅動因素，下載量前十的應用均為 AI 助手。這個羣體的領頭羊是 OpenAI 的 ChatGPT、Google Gemini 和 DeepSeek。僅 ChatGPT 就創造了 34 億美元的全球應用內購買（IAP）收入——這一數字我們在去年底進行了報道。</p>
<p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.10-PM.jpg/query-dz02ODA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.10-PM.jpg/query-dz02ODA"/></p>
<p><strong>圖片來源：</strong> Sensor Tower</p>
<p>在 2025 年，消費者在生成性 AI 應用上花費了 480 億小時，是 2024 年總時間的 3.6 倍，也是 2023 年水平的 10 倍。會話量，即用户打開和使用應用的次數，在 2025 年超過了一萬億次。值得注意的是，這一數字的增長速度超過了下載量，表明現有用户的參與度加深的速度超過了應用新增用户的速度。</p>
<p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.36.58-PM.jpg/query-dz02ODA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.36.58-PM.jpg/query-dz02ODA"/></p>
<p><strong>圖片來源：</strong> Sensor Tower</p>
<p>推動 AI 應用收入和採用的另一個因素是，谷歌、微軟和 X 等大型科技公司正在大力投資其 AI 助手，以挑戰 ChatGPT。在過去一年中，他們以快速的速度推出了新功能，在編碼輔助、內容生成、推理、任務執行、準確性等領域取得了進展。報告特別提到在圖像和視頻生成方面的改進，例如 ChatGPT 在 3 月發佈的 GPT-4o 圖像生成模型，以及谷歌的 Nano Banana。</p>
<p>在頂級 AI 出版商中，OpenAI 和 DeepSeek 佔全球下載量的近 50%，而 2024 年這一比例為 21%。與此同時，大型科技出版商在同一時期的市場份額從 14% 增長到近 30%，擠壓了早期的 ChatGPT 競爭對手，如 Nova、Codeway 和 Chat Smith。</p>
<p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.39-PM.jpg/query-dz02ODA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.39-PM.jpg/query-dz02ODA"/></p>
<p><strong>圖片來源：</strong> Sensor Tower</p>
<p>報告還強調了移動設備在連接用户與生成性 AI 服務中的作用。Sensor Tower 估計，到年底，美國 AI 助手的總受眾超過 2 億，其中超過一半（1.1 億）僅通過移動設備訪問助手。作為對比，2024 年僅有約 1300 萬用户是移動專用。</p>
<p>Techcrunch 活動</p>
<p>舊金山 |2026 年 10 月 13-15 日</p>
<p>除了助手，其他受歡迎的 AI 應用還包括 AI 音樂生成應用 Suno；字節跳動的文本轉視頻應用 Jimeng AI；以及 AI 伴侶應用如 Character.ai 和 PolyBuzz。</p>
<p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.56-PM.jpg/query-dz02ODA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.56-PM.jpg/query-dz02ODA"/></p>
<p><strong>圖片來源：</strong> Sensor Tower</p>
<p>然而，Sensor Tower 發現，AI 並不是去年唯一的收入驅動因素。報告指出，社交媒體、視頻流媒體和生產力等類別的其他應用也推動了增長。例如，消費者在社交媒體應用上的平均每日使用時間為 90 分鐘，總計近 25 萬億小時，同比增長 5%。</p>

OpenAI

<p>在 2025 年，全球消費者在非遊戲移動應用上的支出首次超過了遊戲，達到了約 850 億美元，比前一年增長了 21%。這一增長主要得益於生成式人工智能應用的普及，這些應用的內購收入超過了 50 億美元，下載量翻了一番，達到了 38 億次。主要科技公司在人工智能助手上進行了大量投資，其中 OpenAI 的 ChatGPT 在市場中處於領先地位。總體而言，移動設備在將用户與人工智能服務連接方面發揮了關鍵作用，美國有超過 2 億用户使用人工智能應用</p>

<p>In 2025, consumers spent more money on non-game mobile apps than they did on games for the first time, according to the findings from market intelligence firm Sensor Tower’s annual “State of Mobile” report. While this milestone had been seen in particular markets, like the U.S., or during certain quarters, 2025 marked the first time it occurred globally. Worldwide, consumers spent approximately $85 billion on apps last year, representing a 21% year-over-year increase. The figure was also nearly 2.8x the amount spent just five years ago.</p>
<div class="lb-trans"><p>根據市場情報公司 Sensor Tower 發佈的年度《移動狀態》報告，2025 年消費者在非遊戲移動應用上的支出首次超過了遊戲支出。雖然這一里程碑在某些市場（如美國）或特定季度中曾出現，但 2025 年標誌着這一現象在全球範圍內的首次發生。全球消費者去年在應用上的支出約為 850 億美元，同比增長 21%。這一數字也幾乎是五年前支出的 2.8 倍。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.01-PM.jpg/query-dz02ODA" alt="" original-src="https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.01-PM.jpg?w=680"/></p>
<p><strong>Image Credits:</strong> Sensor Tower</p>
<div class="lb-trans"><p><strong>圖片來源：</strong> Sensor Tower</p>
</div><p>Generative AI, a defining trend over the past year, led the revenue growth, as in-app purchase revenue in this category more than tripled to top $5 billion in 2025. Downloads of AI apps also grew, doubling year-over-year to reach 3.8 billion.</p>
<div class="lb-trans"><p>生成性人工智能作為過去一年中的一個重要趨勢，推動了收入增長，因為這一類別的應用內購買收入在 2025 年超過了 50 億美元，增長了三倍多。AI 應用的下載量也在增長，年同比翻了一番，達到了 38 億次。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.33.06-PM.jpg/query-dz02ODA" alt="" original-src="https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.33.06-PM.jpg?w=680"/></p>
<p><strong>Image Credits:</strong> Sensor Tower</p>
<div class="lb-trans"><p><strong>圖片來源：</strong> Sensor Tower</p>
</div><p>The segment’s growth can be attributed to several factors. For one, the popularity of AI assistants among consumers was a large driver, with all of the top 10 apps by downloads being AI assistants. This group was led by OpenAI’s ChatGPT, Google Gemini, and DeepSeek. ChatGPT alone generated $3.4 billion in global in-app purchase (IAP) revenue — a figure that we reported on late last year.</p>
<div class="lb-trans"><p>該領域的增長可以歸因於多個因素。首先，消費者對 AI 助手的熱情是一個重要驅動因素，下載量前十的應用均為 AI 助手。這個羣體的領頭羊是 OpenAI 的 ChatGPT、Google Gemini 和 DeepSeek。僅 ChatGPT 就創造了 34 億美元的全球應用內購買（IAP）收入——這一數字我們在去年底進行了報道。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.10-PM.jpg/query-dz02ODA" alt="" original-src="https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.10-PM.jpg?w=680"/></p>
<p><strong>Image Credits:</strong> Sensor Tower</p>
<div class="lb-trans"><p><strong>圖片來源：</strong> Sensor Tower</p>
</div><p>In 2025, consumers spent 48 billion hours in generative AI apps, or 3.6x the total time spent in 2024 and 10x the level seen in 2023. Session volume, meaning the number of times users opened and used an app, topped one trillion in 2025. Of note, this figure was growing faster than downloads, suggesting that existing users were deepening their engagement faster than the apps were adding new users.</p>
<div class="lb-trans"><p>在 2025 年，消費者在生成性 AI 應用上花費了 480 億小時，是 2024 年總時間的 3.6 倍，也是 2023 年水平的 10 倍。會話量，即用户打開和使用應用的次數，在 2025 年超過了一萬億次。值得注意的是，這一數字的增長速度超過了下載量，表明現有用户的參與度加深的速度超過了應用新增用户的速度。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.36.58-PM.jpg/query-dz02ODA" alt="" original-src="https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.36.58-PM.jpg?w=680"/></p>
<p><strong>Image Credits:</strong> Sensor Tower</p>
<div class="lb-trans"><p><strong>圖片來源：</strong> Sensor Tower</p>
</div><p>Another factor driving AI app revenue and adoption is that big tech companies like Google, Microsoft, and X have been heavily investing in their AI assistants to challenge ChatGPT. Over the past year, they’ve been rolling out new capabilities at a rapid pace, improving in areas like coding assistance, content generation, reasoning, task execution, accuracy, and more. The report specifically called out improvements in image and video generation, like ChatGPT’s GPT-4o image generation model released in March, and Google’s Nano Banana.</p>
<div class="lb-trans"><p>推動 AI 應用收入和採用的另一個因素是，谷歌、微軟和 X 等大型科技公司正在大力投資其 AI 助手，以挑戰 ChatGPT。在過去一年中，他們以快速的速度推出了新功能，在編碼輔助、內容生成、推理、任務執行、準確性等領域取得了進展。報告特別提到在圖像和視頻生成方面的改進，例如 ChatGPT 在 3 月發佈的 GPT-4o 圖像生成模型，以及谷歌的 Nano Banana。</p>
</div><p>Among the top AI publishers, OpenAI and DeepSeek accounted for nearly 50% of global downloads, up from 21% in 2024. Meanwhile, big tech publishers grew their share of the market from 14% to nearly 30% during this same time, crowding out earlier ChatGPT competitors like Nova, Codeway, and Chat Smith.</p>
<div class="lb-trans"><p>在頂級 AI 出版商中，OpenAI 和 DeepSeek 佔全球下載量的近 50%，而 2024 年這一比例為 21%。與此同時，大型科技出版商在同一時期的市場份額從 14% 增長到近 30%，擠壓了早期的 ChatGPT 競爭對手，如 Nova、Codeway 和 Chat Smith。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.39-PM.jpg/query-dz02ODA" alt="" original-src="https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.37.39-PM.jpg?w=680"/></p>
<p><strong>Image Credits:</strong> Sensor Tower</p>
<div class="lb-trans"><p><strong>圖片來源：</strong> Sensor Tower</p>
</div><p>The report also highlighted the role that mobile plays in connecting users to generative AI services. Sensor Tower estimates that the total audience for AI assistants topped 200 million in the U.S. by year-end, and more than half (110M) were accessing the assistants exclusively on mobile devices. In 2024, for comparison, only around 13 million users were mobile-only.</p>
<div class="lb-trans"><p>報告還強調了移動設備在連接用户與生成性 AI 服務中的作用。Sensor Tower 估計，到年底，美國 AI 助手的總受眾超過 2 億，其中超過一半（1.1 億）僅通過移動設備訪問助手。作為對比，2024 年僅有約 1300 萬用户是移動專用。</p>
</div><p>Techcrunch event</p>
<div class="lb-trans"><p>Techcrunch 活動</p>
</div><p>San Francisco|October 13-15, 2026</p>
<div class="lb-trans"><p>舊金山 |2026 年 10 月 13-15 日</p>
</div><p>Beyond assistants, other popular AI apps included the AI music generation app Suno; ByteDance’s text-to-video app, Jimeng AI; and AI companion apps like Character.ai and PolyBuzz.</p>
<div class="lb-trans"><p>除了助手，其他受歡迎的 AI 應用還包括 AI 音樂生成應用 Suno；字節跳動的文本轉視頻應用 Jimeng AI；以及 AI 伴侶應用如 Character.ai 和 PolyBuzz。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.56-PM.jpg/query-dz02ODA" alt="" original-src="https://techcrunch.com/wp-content/uploads/2026/01/Sensor-Tower-State-of-Mobile-2026-2.31.56-PM.jpg?w=680"/></p>
<p><strong>Image Credits:</strong> Sensor Tower</p>
<div class="lb-trans"><p><strong>圖片來源：</strong> Sensor Tower</p>
</div><p>However, AI wasn’t the only revenue driver last year, Sensor Tower found. Other apps, including those in categories like social media, video streaming, and productivity, also helped fuel the growth, the report noted. For instance, consumers spent an average of 90 minutes per day on social media apps, totaling nearly 2.5 trillion hours, up 5% year-over-year.</p>
<div class="lb-trans"><p>然而，Sensor Tower 發現，AI 並不是去年唯一的收入驅動因素。報告指出，社交媒體、視頻流媒體和生產力等類別的其他應用也推動了增長。例如，消費者在社交媒體應用上的平均每日使用時間為 90 分鐘，總計近 25 萬億小時，同比增長 5%。</p>
</div>

由於 AI 應用的普及，2025 年消費者在移動應用上的支出超過了遊戲