Build with multimodal AI from Google
Gemma 3 is Google's new models for multimodal AI (text, images, video). 1B-27B sizes, 128K context, 140+ languages. Includes ShieldGemma 2 for safety.
