Another day, another proprietary AI decision model with no weights released.
Original headline: Liquid AI Releases d1.2 and d1.2-ultra: Omnimodal Decision Models That Top the AA Decision Index — marktechpost.com
More in AI and machine learning news
Liquid AI has rolled out d1.2 and d1.2-ultra, described as 'omnimodal decision models' that can process text, images, and audio to answer questions. These models don't generate text but instead output calibrated probabilities over a fixed set of answer types, like reading values or locating objects, and they're only available via Liquid AI's hosted API with no open-source weights.
The release is framed around topping the Artificial Analysis Decision Index, with d1.2-ultra scoring a 75, but the practical impact is limited to developers locked into Liquid's ecosystem. The familiar cycle continues: a new version with incremental task additions, a claim of outperforming a proprietary benchmark, and a vague promise to open the weights 'at some point.' For most engineers, this is just vendor noise until the models are actually accessible.
Read the original at marktechpost.com
Satirical commentary on real tech news, summarised with AI. Read the original source for the facts. How ZeroDaily works
