Liquid AI ships d1.2 and d1.2-ultra omnimodal decision models, tops Decision Index at 75

calabi_and_yau · x · 2026-11-16

Liquid AI released d1.2 and d1.2-ultra, self-described state-of-the-art omnimodal decision models that process text, images, and audio with new structured output capabilities, claiming #1 on Artificial Analysis' Decision Index (75). New outputs include reading values from inputs, finding all text spans of a kind, locating objects in images (single or all-instance boxes), estimating numbers with ranges, and multi-select choices.

Related event: Liquid AI Releases d1.2 Series, Topping Decision Index(4 posts)→

Original post →

More from Models

Models channel →