Google is expanding its AI Mode in Search to five additional languages: Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese. After piloting the experience in English for months, the company says the multilingual rollout will let more users ask complex questions and get AI-generated overviews in their native tongue.
Five new languages widen access
Hema Budaraju, VP of Product Management for Google Search, said the expansion is designed to make AI Mode feel native, not translated. Hindi, Japanese, and Korean join Indonesian and Brazilian Portuguese—five languages with deep cultural nuance and distinct writing systems that are often underserved by cutting-edge AI features at launch.
AI Mode sits alongside traditional results, accessible via a dedicated tab and a button in the search bar. Google has been testing the experience in more markets and hinted it could become the default for some users over time; Google DeepMind’s Logan Kilpatrick recently indicated the company is moving in that direction.
Why these languages matter
The five additions collectively represent hundreds of millions of speakers and some of the world’s most engaged mobile internet populations. India has more than 750 million internet users, according to IAMAI–Kantar, and Hindi remains a dominant language across search and social. Japan counts roughly 125 million native Japanese speakers; South Korea has around 80 million Korean speakers globally; Indonesian is used by nearly 200 million people as a first or second language; and Brazil’s Portuguese-speaking community exceeds 210 million people.
Multilingual support is also a competitive necessity. While English still accounts for roughly half of web content, estimates from W3Techs show sizable shares in Japanese and Portuguese, and user growth is fastest in regions where English is not the primary language. Bringing AI Mode to these markets positions Google to defend share against AI-first search challengers that already localize aggressively.
What AI Mode can do now
AI Mode is powered by a tailored version of Gemini 2.5, Google’s multimodal model that can reason over text and other inputs. In practice, that means a user in Tokyo can ask in Japanese for “late-night ramen near me that takes card and has vegetarian sides,” and receive an overview synthesizing reviews, menus, and hours, alongside links for deeper reading. A Korean speaker might request a step-by-step plan to contest a parking ticket, then refine the instructions with follow-up questions, all in Korean.
Google is also layering in agentic capabilities—early features that take action on your behalf. In select tests, AI Mode can help find restaurant reservations, with plans to add local service appointments and event ticketing. These capabilities currently sit behind Google’s premium experiments and are limited to U.S.-based Ultra subscribers as the company validates quality and safety controls.
Availability, pricing, and where it appears
AI Mode first appeared as an experiment for Google One AI Premium customers before broadening in English to more markets. The new language support is rolling out in stages, and Google typically staggers access through Search Labs toggles and regional eligibility.
The base experience is included with Search when enabled, surfaced via the AI Mode tab or prompt button. Advanced agentic features are part of a paid tier for now; the Ultra plan is priced at $249.99 per month in the U.S. as Google tests higher-cost infrastructure and more proactive tasks.
Implications for search and publishers
AI-generated overviews can change how people click. Publishers and e-commerce sites worry that instant answers reduce referral traffic, particularly on short, fact-seeking queries. Google has pushed back, saying AI Mode and AI Overviews aim to direct users to authoritative sources and are not designed to suppress clicks. Independent analysts note the impact likely varies by query type: navigational and purchase-intent searches tend to continue driving visits, while some informational queries consolidate into overviews.
For users, the multilingual expansion should improve clarity and confidence. Asking nuanced, multi-part questions in one’s own language—about visas, health guidance, or local services—reduces translation friction and the risk of misunderstanding. For Google, it’s a stress test of Gemini 2.5’s reasoning across scripts, honorifics, and regional idioms, where context errors can be costly.
What to watch next
Key signals to monitor include the speed of rollout across these languages, whether AI Mode becomes the default for more users, and how quickly agentic tasks expand beyond the U.S. If Google can keep quality high—and demonstrate meaningful click-through to trusted sources—the multilingual push could cement AI Mode as a core part of everyday search for hundreds of millions of people.