Sign in
← AI PulseAug 12, 2026

Wire · news · Single-source brief

Google DeepMind Introduces SL2T for Sign Language AI in Consumer Products

Google DeepMind has launched a massively multilingual sign-language-to-text (SL2T) translation model, bringing sign language AI into consumer products like Gboard and Live Transcribe on Pixel 11, starting with American Sign Language (ASL) to English.

By Illumora Editorial

Source · Aug 12, 2026, 2:01 PM · On Illumora · Aug 12, 2026, 2:07 PM

Video from the primary source — playable here so you can stay on Illumora.
Also from the source · 2 of 4
Also from the source · 3 of 4
Also from the source · 4 of 4

Rewritten from one allowlisted primary — not independent enterprise reporting. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →Google DeepMind Blog — Putting sign language AI into users’ hands
Save

Google DeepMind has introduced sign-language-to-text (SL2T), a new model designed to power sign language features for Deaf and hard of hearing users. This development aims to extend the benefits of AI-driven language processing, which has rapidly advanced for spoken languages, to the more than 200 sign languages used by an estimated 70 million Deaf and hard of hearing individuals globally.

The SL2T model is described as a breakthrough in quality and generality. It is being integrated into consumer products for the first time, enabling sign-to-text dictation in Gboard and Live Transcribe on Pixel 11, initially supporting American Sign Language (ASL) to English. Google DeepMind states that more devices and additional languages will follow.

Key Points

  • SL2T is a massively multilingual sign-language-to-text translation model.
  • It powers sign-to-text dictation in Gboard and Live Transcribe on Pixel 11.
  • The initial release supports American Sign Language (ASL) to English translation.
  • SL2T processes sign language inputs as pose landmark locations, not raw camera feeds.
  • An on-device model, MediaPipe Holistic, tracks signer points, sending only geometric coordinates for translation.
  • The model is trained on over 100,000 hours of data across more than 50 sign languages.
  • SL2T achieves a zero-shot score of 70 BLEURT on the FLEURS-ASL (sd-test) benchmark.

Context

According to Google DeepMind, while AI has advanced rapidly for spoken languages, this technological progress has not reached sign languages. Sign languages are independent natural languages with distinct grammars and lexicons, requiring true machine translation rather than sequential sign-to-word transformations. Additionally, accurately tracking the simultaneous movements of hands, arms, torso, head, and face at high frame rates presents a complex and computationally demanding computer vision challenge.

Why It Matters

This development offers Deaf users a new way to interact with technology, enabling them to sign to their phones for tasks like web searches, drafting messages, or interacting with AI assistants. It also opens possibilities for bridging communication gaps between Deaf and hearing communities by providing sign language processing capabilities similar to those available for spoken languages.

What To Do

  • Note the availability of sign-to-text dictation in Gboard and Live Transcribe on Pixel 11.
  • Watch for future updates regarding additional devices and sign languages supported by SL2T.
  • Observe how the MediaPipe Holistic model is used for on-device pose tracking to protect user privacy.
  • Consider the implications of direct translation from pose landmarks, bypassing intermediate glosses, for improving translation quality.