Google · Gemma 3
Google's 27B Gemma 3 multimodal instruct model — strongest of the Gemma 3 family with vision understanding and 128K context.
Gemma-3-27b-it Prompt — rank #69
