Gemini 3.5 Live Translate launches with 70+ language support
Google DeepMind has released Gemini 3.5 Live Translate, an audio model delivering near real-time speech-to-speech translation across more than 70 languages.
Score breakdown
Gemini 3.5 Live Translate extends Google's translation capabilities from text to live speech-to-speech audio, covering over 70 languages in near real-time.
- 01Gemini 3.5 Live Translate is Google DeepMind's latest audio model for live speech-to-speech translation.
- 02The model supports over 70 languages.
- 03It delivers near real-time translation.
Google DeepMind has released Gemini 3.5 Live Translate, its latest audio model aimed at fluid, natural-sounding live speech-to-speech translation. According to the post, the model delivers near real-time translation in over 70 languages. The authors of the post are identified as Anuda Weerasinghe (Product Manager) and Tony Lu (Senior Staff Software Engineer).
The release is positioned as a milestone in Google's translation history, which the post traces back roughly twenty years to early machine learning experiments.
The release is positioned as a milestone in Google's translation history, which the post traces back roughly twenty years to early machine learning experiments. Google's translation services currently process over a trillion words per month for billions of users across its products. The source text is truncated before further technical or product details are provided.
Key facts
- 01Gemini 3.5 Live Translate is Google DeepMind's latest audio model for live speech-to-speech translation.
- 02The model supports over 70 languages.
- 03It delivers near real-time translation.
- 04The post is authored by Anuda Weerasinghe (Product Manager) and Tony Lu (Senior Staff Software Engineer).
- 05Google's translation services process over a trillion words per month for billions of users across its products.
- 06Google's machine translation work began approximately twenty years ago.
Topics
Summary and scoring are generated automatically from the original article. We always link back to the publisher and never republish images or paywalled content. Last processed Jun 9, 2026 · 17:05 UTC. How this works →