Google Takes Dreambeans to India With 13 Local Languages

Google launched its experimental AI app Dreambeans in India on Monday, the app's first market outside the U.S., with support for 13 local languages including Hindi, Punjabi, Tamil, Telugu, and Bengali, alongside English. The move targets the world's second-largest internet market less than two months after Google opened the app to all eligible U.S. users aged 18 and older.

Dreambeans, available on Android and iOS, doesn't wait for a prompt. It pulls data from connected Google services — Calendar, Gemini, Gmail, Photos, Search, and YouTube — to proactively generate personalized, illustrated stories about users' interests and everyday lives. Google introduced the app in June with static AI-generated illustrations, then recently added a feature using its Veo video-generation model to animate images in select stories.

Why India first? Google said Indian users showed strong interest in gaining access, and the company's services are widely used there. India also stood out for its high overlap of users across Gemini, Gmail, and Google Photos, Google product manager Gozde Oznur said during a press briefing. For a product that depends on users already living inside the Google ecosystem, that overlap is the fuel.

The U.S. numbers explain the urgency. Dreambeans recorded nearly 55,000 downloads in the U.S. since its June debut, according to estimates from market intelligence firm Sensor Tower shared with TechCrunch. About 63% of those downloads came in September, when Google opened the app to all eligible U.S. users aged 18 and older. Then downloads dropped 84% month over month to about 5,500 in October.

The competitive gap is stark. In the first 30 days after opening to all eligible U.S. users, Dreambeans recorded an estimated 29,000 downloads. During their respective first 30 days after launch, Meta's Muse racked up 7.4 million, ChatGPT 2.8 million, Grok Bot 362,000, and Claude 155,000, Sensor Tower estimates show. Google's counter is positioning: unlike those general-purpose AI apps, Dreambeans generates personalized stories rather than answering a broad range of queries.

What does the language limit change for Indian users?

Plenty. Dreambeans supports many widely spoken Indian languages, but users can select only one language at a time for their generated stories, Oznur said. The app can draw on information across Google services in multiple languages, but it cannot currently generate stories that mix languages.

That's a real constraint in India, where many people regularly switch between languages — often mixing Hindi and English in what's commonly known as Hinglish. An app that can't follow a user's natural code-switching risks feeling out of step with how the country actually talks.

What about privacy and sensitive data?

Dreambeans' dependence on users' Google accounts raises questions about data access. The app uses Google's Personal Intelligence technology to tap connected services, which can contain sensitive personal information some users may not want to share with an AI app just to get personalized content.

Oznur said users can choose which Google services to connect and disconnect them at any time, and can delete their Dreambeans data from within the app. She asserted that information deleted from underlying connected services is also removed from Dreambeans' personalization system. Users can also tell the app to avoid certain subjects, including business-related information. Dreambeans is available only to personal Gmail account holders, not enterprise users, and generated stories are visible only to the account holder.

The proactive model carries another risk: surfacing memories users may find upsetting. Oznur acknowledged Dreambeans could generate stories involving people who have died. She compared the controls — which let users stop the app showing certain subjects — to features in Google Photos that limit which memories appear.

The next step, Oznur told reporters, is making Dreambeans more actionable: moving beyond personalized stories and recommendations to help users act on the app's suggestions. For practitioners, that points toward Google turning passive content generation into a task-execution surface built on the deepest behavioral dataset it already owns.