Leading European financial services organizations embrace open models to build proprietary intelligence
European banks, payment firms and financial data providers use NVIDIA GPU computing and open model weights, often on their own infrastructure, for transactions, speech, research and fraud detection. Revolut, bunq and Stripe are among the examples cited.
What the article says
In an October 21, 2026 post tied to GTC Berlin, NVIDIA describes how European financial institutions are using open-weight models on its accelerated computing, often on their own infrastructure. Revolut is rebuilding its stack around PRAGMA, a transaction foundation model serving an initial 80 million customers; Starling Bank trained an agentic model on U.K. sovereign cloud, and Lloyds built ATUM, a transaction foundation model.
Other examples: bunq says its Finn assistant on open models matches leading closed models at 60% of the cost per answer; LSEG aims for up to 10x faster analysis of tick data; Stripe cites 42x faster pipelines in offline testing. The post frames open models on controlled infrastructure as a way to build proprietary intelligence. The figures are vendor-reported and describe specific deployments.
Summary written by AnnuIA in its own words from the official page of NVIDIA. The content and rights belong to the publisher: please read the original article.
More announcements from NVIDIA
- Agentic AI joins the search for extraterrestrial signals
- Thin, light and built for everything: NVIDIA RTX Spark Windows laptops gain new models, agent tools and games
- How AI is changing the pace of discovery for millions living with dementia
- NVIDIA Cosmos opens new tools to train, evaluate and deploy world models