# 影像生成功能已成為 AI 行動應用成長的最大推手。

*genai · news · 2026-05-05 · Firstpost*

## Key points

- 影像生成功能現在推動的 AI 應用下載量是對話更新的 6.5 倍。
- Google Gemini 的 Nano Banana 影像模型引發下載量四倍增長，但營收成長有限。
- 只有 OpenAI 將影像生成的成功轉化為收益，估計在發布後 28 天內賺取 7,000 萬美元。
- 大多數用戶嘗試影像 AI 功能卻未轉換為付費訂閱，凸顯變現挑戰。
- DeepSeek R1 主要因其低成本訓練方法吸引了 2,800 萬次下載。

Not long ago, the appeal of artificial intelligence apps lay in conversation. Users flocked to download tools that could answer questions, draft emails, or simulate human-like dialogue. But the centre of gravity is shifting. Today, it is not what AI says, but what it creates, that is pulling users in. A new report from Appfigures suggests that image-generation features have become the biggest driver of growth for AI mobile applications, outpacing traditional model upgrades by a wide margin. According to the firm’s analysis, releases of image-based AI models are generating 6.5 times more downloads than standard updates focused on conversational improvements. Image AI models driving downloads across ChatGPT, Gemini and others The data points to a clear change in user behaviour. Earlier waves of growth were tied to improvements in chatbot capabilities or the addition of features such as voice interaction. Now, visually creative tools appear to be the main attraction. Apps like ChatGPT and Google Gemini have both seen tens of millions of new downloads following the rollout of image-generation models. These launches gave users a tangible reason to try out the apps, offering the ability to generate images from text prompts, edit visuals, or experiment with creative outputs. One standout example is Google’s Gemini. The introduction of its image model, Nano Banana, alongside the Gemini 2.5 Flash update last August, led to a surge of more than 22 million additional downloads within just 28 days. That spike represented a more than fourfold increase compared to typical download levels over a similar period, according to the report. The pattern highlights how visual features are reshaping the competitive landscape. Unlike incremental improvements in chatbot performance, image tools offer immediate, shareable results that can quickly capture user attention and spread across social platforms. AI app revenue vs downloads: Why growth doesn’t always mean profits However, the surge in downloads does not necessarily translate into financial success. The Appfigures report cautions that while image model launches are effective at attracting users, converting that interest into paying subscribers remains a challenge. In Gemini’s case, despite the significant spike in downloads driven by Nano Banana, the app generated only an estimated $181,000 in gross consumer spending during the same 28-day window. This suggests that many users are experimenting with the technology without committing to paid tiers. A similar trend was observed with Meta’s AI efforts. Its Vibes feature also boosted downloads but failed to deliver meaningful revenue gains. Among the major players, only OpenAI appears to have successfully turned attention into income. The launch of its 4o image-generation model led to an estimated $70 million in gross consumer spending over 28 days, significantly outperforming competitors in monetisation. The report also examined DeepSeek R1, though its trajectory differed. The model drove 28 million downloads following its release in January 2025, but this surge was less about image capabilities and more about curiosity. DeepSeek’s sudden rise, fuelled by revelations about its low-cost training methods, turned it into an overnight sensation and drew users eager to see what set it apart. Taken together, the findings point to a broader shift in the AI app economy. Visual creativity is emerging as the primary hook for user growth, even as monetisation strategies struggle to keep pace. For developers, the challenge now is not just to attract users with eye-catching features, but to persuade them that these tools are worth paying for.

**Companies:** Google, OpenAI, Meta
**Countries:** United States

[Read the full story on Firstpost](https://www.firstpost.com/tech/ai-image-generation-tools-boost-app-installs-surpassing-chatgpt-gemini-chatbot-updates-14007657.html)

---

Canonical: https://newsio.io/zh-TW/n/c357c11a-e2b7-48d6-9656-c51545ff0393/ai-ai-6-5
Summarized by Newsio from Firstpost. https://newsio.io/how-it-works