Google launches Dreambeans in India to create AI stories from personal data

Expanding to India

Google has officially launched Dreambeans in India, marking the first time the company has brought the experimental app to a market outside of the United States. The move is intended to reach the world's second-largest internet market by offering personalized visual stories to a wider audience. The app is currently available for users on both Android and iOS devices.

The expansion includes support for 13 local languages, providing accessibility for speakers of Hindi, Punjabi, Tamil, Telugu, and Bengali, in addition to English. Google selected India as its first international destination due to the high level of interest from local users and the widespread use of its existing ecosystem. The company specifically noted the high overlap of users who already utilize Gemini, Gmail, and Google Photos.

While the app has seen a presence in the U.S. since June, its performance there has been varied. Market intelligence from Sensor Tower suggests that while downloads spiked in September, they saw a significant drop in October. Despite these fluctuations, Google is betting that the Indian market will provide a robust environment for testing how people interact with AI-generated life summaries.

How the technology works

Dreambeans operates differently than general-purpose chatbots because it does not require users to enter specific prompts to receive content. Instead, it uses Google’s Personal Intelligence technology to proactively generate illustrated stories based on a user's digital activity. It pulls information from a variety of connected services, including Calendar, Gemini, Gmail, Photos, Search, and YouTube.

The app creates these stories by analyzing interests and everyday life events found across these various platforms. While the app initially only featured static AI-generated illustrations, Google recently integrated its Veo video-generation model. This update allows the app to animate images within select stories, providing a more dynamic visual experience for the user.

A notable technical limitation currently exists regarding language processing in the Indian context. Although the app supports many local languages, users can only select one language at a time for their stories. The system cannot currently generate stories that mix languages, which may impact its appeal for people who frequently use

U.S. launch and market competition

The app's initial rollout in the United States provided a baseline for how it compares to other major AI tools. During the first 30 days after opening to all eligible U.S. users aged 18 and older, Dreambeans recorded an estimated 29,000 downloads. This figure was significantly lower than the debut numbers for several prominent general-purpose AI applications, such as Meta’s Muse, ChatGPT, Grok Bot, and Claude.

Because Dreambeans is designed for specific personalized storytelling rather than answering a broad range of general queries, its growth pattern differs from these competitors. While it did not match the massive scale of the larger models, it serves a unique niche in the market. Google is now looking to see if the Indian market will offer a different trajectory for the app's adoption and growth.

Recent download trends and future goals

Data from the U.S. market shows that the app experienced a significant surge in interest during September, which accounted for about 63% of the total downloads since the app's June debut. However, this momentum shifted quickly in the following month, with downloads dropping 84% month over month to approximately 5,500 in October. These fluctuations highlight the experimental nature of the product as Google refines the user experience.

Looking ahead, Google is not content with just providing passive content like stories and recommendations. Product manager Gozde Oznur told reporters that the company is exploring ways to make the app more actionable. The goal is to move beyond simple storytelling to a system that helps users actually take steps or perform tasks based on the suggestions the AI generates.