New MAI family includes reasoning, coding, image, voice and transcription models aimed at developers and enterprises.
Google has launched Gemma 4 12B, a new open-source multimodal AI model that supports text, image, and native audio inputs ...