Gemini 3.5 Transcribe: Google's Most Precise Speech AI Yet
Gemini 3.5 Transcribe is Google's new speech-to-text model, live in public preview with 2.6% word error rate and support for more than 85 languages.
6 articles
Gemini 3.5 Transcribe is Google's new speech-to-text model, live in public preview with 2.6% word error rate and support for more than 85 languages.
Google's Gemini watermark toggle removes the visible AI mark from Nano Banana, Omni, and Lyria outputs. The EU is excluded and 3 countries need AI Ultra.
Google Vids adds AI avatars, Veo 3.1 video generation, Lyria 3 music, and YouTube export. Free users get 10 videos/month — see how it rivals HeyGen.
Google launches Lyria 3 in Gemini, a free AI music generator creating 30-second tracks from text, images, and video. Available globally for users 18+.
Google Flow, powered by Veo 3.1, is now available for Workspace Business, Enterprise, and Education users. Create 8-second video clips with audio.
Google Veo 3.1 adds native 9:16 vertical output for YouTube Shorts, 4K upscaling, and better character consistency. Available now in the Gemini app.