Nace AI Releases Financial Decision Model Drex for Finance
Nace AI has released Drex for Finance, a multimodal decision model for financial scenarios, used for sub-second financial decisions such as fraud detection; it can combine documents or transaction logs with a fixed set of candidate answers to output a probability score for each answer. On 3,361 decisions across 48 held-out financial benchmarks not used in training, its accuracy was 85.7%, compared with 70.9% for GPT-6 Luna. The model is available only in Nace Console.
Why read this
AI score 56
The original text provides the accuracy of Drex for Finance on 48 financial benchmarks and its comparison with GPT-6 Luna, which can be used to assess the positioning of specialized decision models.
Topics
View original postAnnouncing Drex for Finance - the multimodal decision model, post-trained for sub-second financial decisions. Financial operations require precise, regulatory-grade decisions with zero hallucinations across filings, support, credit risk, live fraud prevention, quant signals, and compliance. Evaluated on 48 held-out finance benchmarks: - 85.7% accuracy across 3,361 decisions (+14.8% vs. GPT-6 Luna) - 77.3 Finance Index (+13.7 vs. Jev 1.14, +11.5 vs. Drex-Max v1.1) Trained for sub-second financial intelligence across the entire pipeline: - Financial Documents & Filings: Extraction decisions, classification, and clustering - Credit & Lending: Automated underwriting, PD modeling, and EWS - Fraud & AML: Sub-second transaction scoring, ATO defense, and SAR clustering - Markets & Trading: Regime detection, meta-labeling, and LOB execution signals - Insurtech & RegTech: Claims fraud detection, sanctions screening, and audit reconciliation Available exclusively on Nace Console: https://t.co/vdEavE1LhL
View original
Announcing Drex for Finance - the multimodal decision model, post-trained for sub-second financial decisions. Financial operations require precise, regulatory-grade decisions with zero hallucinations across filings, support, credit risk, live fraud prevention, quant signals, and compliance. Evaluated on 48 held-out finance benchmarks: - 85.7% accuracy across 3,361 decisions (+14.8% vs. GPT-6 Luna) - 77.3 Finance Index (+13.7 vs. Jev 1.14, +11.5 vs. Drex-Max v1.1) Trained for sub-second financial intelligence across the entire pipeline: - Financial Documents & Filings: Extraction decisions, classification, and clustering - Credit & Lending: Automated underwriting, PD modeling, and EWS - Fraud & AML: Sub-second transaction scoring, ATO defense, and SAR clustering - Markets & Trading: Regime detection, meta-labeling, and LOB execution signals - Insurtech & RegTech: Claims fraud detection, sanctions screening, and audit reconciliation Available exclusively on Nace Console: https://t.co/vdEavE1LhL
Body
Nace AI has launched Drex for Finance, a model designed for split-second financial tasks, like determining whether a card payment looks fraudulent.
It can use a document or a transaction log along with a fixed set of possible answers, then deliver a probability score for each one.
In tests with 3,361 decisions from 48 finance benchmarks not included in training, it was right 85.7% of the time, compared to 70.9% for GPT-6 Luna.