LaunchAI ModelsNovember 6, 2026

Ant launches Ling-3.2-flash multimodal model with million-token context

Read original source →technode.com

Ling-3.2-flash has 560B total parameters but activates only 25B per inference step, accepting text, images, and video. Ant targets complex visual understanding, multistep agent tasks, 3D interaction, and medical multimodal use.

1 source

More stories today

Open the live feed