Tag
Modal announces day 0 support for the Step 3.7 Flash AI model, a 198B parameter MoE with 11B active parameters, 256K context, three reasoning levels, and native image and video understanding.