The goal of
@Android has always been to bring the best of computing to everyone.
For years, speech-to-text has made mobile communication effortless. And although many in the Deaf and hard-of-hearing community prefer using sign language, typing in English on a keyboard has remained the default requirement on smartphones.
With the launch of the new Pixel 11 series, we introduced sign-to-text translation, powered by
@GoogleDeepMind’s multilingual sign language-to-text (SL2T) translation model.
Built in deep partnership with the Deaf community and our AI Sign Language Advisory Committee (AISLAC) from day one, sign-to-text translates American Sign Language into written English in real time across Gboard and Live Transcribe. Whether you're drafting a message, prompting Gemini, or ordering coffee, you can communicate naturally in your own language. Using your device’s front camera, the sign-to-text feature tracks upper-body, hand, and facial movements in real time. Then it sends a 2D map of your movements to the server, not your video, to be translated into English.
It’s been awesome seeing the early love and reactions over the past couple of days, from folks testing it out at the Google Store to seeing the excitement from Deaf users and community members imagining what this future looks like.
It’s still early days, and we’re committed to continuously improving the experience as we bring sign-to-text to more Android devices and support more global sign languages over time.
Incredible work by the teams across Android and DeepMind, and a huge thank you to our community partners for guiding this important milestone!
See how this feature works in a demo by Daniel Durant, the acclaimed Broadway star 👇