Multilingual voice assistants: from English-first to global by design
A multilingual voice assistant is an AI system that can understand, respond to, and smoothly switch between multiple spoken languages in a single conversation without forcing users to change any settings or stick to one fixed language. Until recently, most AI assistants treated English as the default and pushed everyone else toward clumsy workarounds like manual language toggles or typing in their preferred language. Google’s Gemini app and Anthropic’s Claude Voice Mode are now shifting that model. Both are adding richer language options and new ways to control how you speak to them, so people who naturally code-switch can talk the way they do with friends and family. These changes are not only about convenience; they push AI assistant accessibility toward a world where speaking more than one language is a normal use case rather than an edge case.
Gemini voice commands now understand over 70 languages
Google’s latest update turns the Gemini mic into a powerful multilingual voice assistant. The app’s voice input now supports “over 70 languages,” according to a post shared by Josh Woodward, Google’s VP for Gemini. Crucially, users do not need to adjust language settings before speaking. You can start in one language, switch to another mid-sentence, and Gemini voice commands still work. One test described in the source shows a single spoken request mixing Hindi and English, which Gemini handled without confusion. This upgrade is rolling out on Android and iOS, with web support expected “in another week or so.” For people who balance multiple languages in everyday life, this feels closer to natural conversation: fewer menus, fewer taps, and less mental friction between what you want to say and what the assistant can understand.
Claude Voice Mode adds multilingual options and Push-to-talk
Anthropic’s Claude Voice Mode is also undergoing a significant upgrade that moves it beyond English-first design. Users are starting to see a new Voice settings screen in the mobile apps with a long list of voice input languages, including Spanish (Latin America), German, Portuguese, Chinese, Japanese, Russian, and Ukrainian. The app now offers two main interaction styles: a hands-free mode for continuous, natural back-and-forth conversations, and a Push-to-talk mode where you press and hold while speaking, then release to send your message. One early user said the new experience is “100 times better than ChatGPT,” a subjective claim but a sign of how much the interface has changed. The rollout is gradual and availability varies, so some accounts show the new options while others are still on the older interface as Anthropic tests and expands the features.
New controls, code-switching, and the next wave of voice-first AI
Taken together, these updates signal a broader shift in AI assistant accessibility. Gemini’s automatic detection of over 70 voice input languages, plus its ability to process mixed-language commands, means code-switching is not treated as a problem to fix but as a normal way of speaking. Claude’s Voice Mode, meanwhile, pairs multilingual support with interaction flexibility: hands-free for long conversations, Push-to-talk when you want tighter control over what gets sent. Both apps are rolling features out in stages, which explains why some users see the new interfaces before others. A mysterious phone-call-style icon spotted in Claude’s iOS build hints that Anthropic may be experimenting with more phone-like voice sessions. While the details are still unclear, the direction is obvious: voice assistants are becoming more conversational, more flexible, and finally more aligned with how multilingual users speak in real life.
