Multimodal decision model post-trained for sub-second, regulatory-grade financial decisions across filings, credit risk, fraud prevention, trading signals, and compliance. It scores 85.7% accuracy on 3,361 decisions from 48 held-out finance benchmarks, outperforming GPT-6 Luna, and is available exclusively via the Nace Console.
Drex for Finance is Nace.AI’s decision model for financial workflows. It takes case information and a fixed set of answer options, then returns scores or probabilities for those options. Nace lists uses including invoice and filing checks, credit-risk classification, fraud and dispute triage, market-signal classification, insurance-claim review, and banking-request routing. The model is accessed through Nace Console using the model ID `drex-max-finance`; Nace also documents Python and TypeScript SDKs and an API. Nace reports benchmark results on 48 public finance test sets, while those figures are publisher-reported.
