Affiliate links on Android Authority may earn us a commission. Learn more.
Gemini Live will soon be able to answer questions about things in your camera viewfinder
- Google will soon integrate Project Astra into Gemini Live, allowing users to have a conversation with the AI chatbot about things in their camera viewfinder.
- The company says that the integration will help users with things like solving a calculus problem or assembling furniture by simply sharing their camera feed with Gemini.
- The upcoming AI experience will also get Messages and Calendar app integration, which will let users perform various actions during the conversation.
Google gave us an early look at Project Astra at I/O this May with demos of an AI chatbot answering questions about things in the camera viewfinder. At the time, the company said that some of its capabilities would make it to the Gemini app later this year, and it has now unveiled a few more details.
Towards the end of today’s Made by Google event, Google said that it would soon integrate Project Astra into Gemini Live — the AI chatbot’s new conversational mode. The integration will allow users to have a conversation with Gemini about things in their camera viewfinder and get answers in real-time. For example, it will be able to offer solutions to a calculus problem or help with furniture assembly.
In addition, the new experience will support a few Google apps, like Messages and Calendar, allowing users to send messages or check their schedule during the conversation. Although you’ll have to wait a few months to try the Project Astra integration, Gemini Live is already rolling out.
The new voice chat mode is exclusively available to Gemini Advanced users on Android and it currently supports ten voices. The mode supports conversations in English at the moment, but Google plans to add more languages and extend the feature to iOS devices in the coming weeks.
It’s also worth noting that Google’s own disclaimer for Project Astra/Gemini Live noted that the product demo was for “aspirational purposes.” So there’s every chance the real-world feature will be much slower and less accurate.