Gemini Live will soon offer recipes in 40 different languages

Gemini Live began as an English-language chat AI, but has been taking language courses. As a result, it will soon offer support for more than 40 languages. The real-time chat feature for users of Android devices promises to connect a significantly larger portion of the world’s population to the service this way.

It is possible to install up to two languages simultaneously on one device. This is useful for multilingual users and users moving in international or multicultural contexts. Users can easily set their preferred languages through the Google app. This is done somewhat anachronistically through the settings for Google Assistant and then choosing ‘Languages.’

Gemini now integrates better with other apps, so switching between them is not –or at least a lot less– necessary. Think of Gmail, Maps, YouTube, Calendar and Tasks. Want to scan a concert flyer and check if you’re free that day? Then Gemini checks that for you via Google Calendar. Have you been emailed a recipe? Then you can ask Gemini to add the ingredients directly to your shopping list in Keep.

Sponsored
young moneyJeden Freitagmorgen umsetzbare und persönliche Einblicke in Vermögensaufbau, Steueroptimierung und mehr - ohne fragwürdige Interessen von Maklern und Finanzjargon.

Good memory

Gemini Live can remember your conversations, allowing you to pause a conversation in between, change the subject or go deeper if you wish. The AI can provide users with advice or help plan events or create a task list. Ten different voices are available.

Users who feel bored, alone or just curious can use Gemini to search for interesting historical events, local topics or themes that match previously expressed interests. A bit similar to the familiar Wikipedia rabbit hole but with the help of AI.

It was previously necessary to have a subscription to Gemini Advanced to use Gemini Live. That requirement has been quietly dropped, although Gemini Live does not use Gemini Pro, unlike Gemini Advanced. For on-device processing, Google provides the Nano model. Gemini’s lightest multimodal model overall is Flash.

Good news for users of Gemini Advanced: Google has rolled out version 1.5 of this LLM, which promises to be faster and more powerful. This update should provide better performance for complex conversations and mathematical queries. Whereas Gemini used to sometimes struggle with stacked instructions, the new version now understands what you mean better and responds faster.

Competition makes itself heard

This improvement comes at a time when competitor ChatGPT has also launched a new model, the o1 preview, based on the Strawberry model. It’s going to be interesting to see how the performance of the new Gemini Pro compares to this advanced model from ChatGPT, which, by the way, is also paid.

Microsoft is also launching live chat via Copilot, free and even available on iOS devices. It is worth remembering that not all features offered by these different vendors are available everywhere at once, nor are they always available in every language.

Sponsored
Livelong NewsletterThe free weekly newsletter on longevity breakthroughs that will help you look good, feel better and live longer - join 40,000+ readers.

Reply

or to participate.