Imagine inventing attention and getting the runner up test of time award 😂 🥈 Congrats to all first and second place winners!

Real-time speech was so two months ago

Happy to see the first feature powered by Gemini native audio outputs ship out to public - especially since it's MASSIVELY multilingual. Lots more coming soon 😉

New multilingual speech restoration paper out Miipher-2 🚀! The RTF on a TPU is 0.0078: 1 million hours of data can be cleaned in 3 days using just 100 TPUs! Paper: arxiv.org/abs/2505.04457 Demo: google.github.io/df-conformer/m…

iykyk

🎧 Audio Overviews are now available in 45 languages. Learn on-the-go by turning your documents, slides, Deep Research reports and more into podcast-style conversations.

Google silently launched a way to describe a speaker's voice "style" in plain English. I've never seen anything like it. It seamlessly does different accents, dialects, tone, and languages. Check out Chinese vs Jamaican and Tamil/English vs Punjabi/English in this clip below!

!!!